data_analysis/Transactions dataset research task.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import datetime\n",
    "import matplotlib.pyplot as plt\n",
    "from scipy import stats as st\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = pd.read_csv('transactions_dataset.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "data['order_date'] = pd.to_datetime(data['order_date'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 534,
   "metadata": {},
   "outputs": [],
   "source": [
    "del data['Unnamed: 0']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 536,
   "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>order_id</th>\n",
       "      <th>customer_id</th>\n",
       "      <th>order_date</th>\n",
       "      <th>is_canceled</th>\n",
       "      <th>number_of_paid_orders_before</th>\n",
       "      <th>date_minus_time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>52285</td>\n",
       "      <td>2016-01-01 00:09:51</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2016-01-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>58296</td>\n",
       "      <td>2016-01-01 14:36:14</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2016-01-01 14:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>30609</td>\n",
       "      <td>2016-01-01 14:47:46</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2016-01-01 14:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>30256</td>\n",
       "      <td>2016-01-01 15:06:36</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2016-01-01 15:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>52642</td>\n",
       "      <td>2016-01-01 16:27:37</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>2016-01-01 16:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>5</td>\n",
       "      <td>56257</td>\n",
       "      <td>2016-01-01 17:14:02</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>2016-01-01 17:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>6</td>\n",
       "      <td>9948</td>\n",
       "      <td>2016-01-01 18:45:36</td>\n",
       "      <td>0</td>\n",
       "      <td>176</td>\n",
       "      <td>2016-01-01 18:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>7</td>\n",
       "      <td>39536</td>\n",
       "      <td>2016-01-01 19:17:55</td>\n",
       "      <td>0</td>\n",
       "      <td>11</td>\n",
       "      <td>2016-01-01 19:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>8</td>\n",
       "      <td>26729</td>\n",
       "      <td>2016-01-01 19:47:50</td>\n",
       "      <td>1</td>\n",
       "      <td>72</td>\n",
       "      <td>2016-01-01 19:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>9</td>\n",
       "      <td>26729</td>\n",
       "      <td>2016-01-01 20:01:01</td>\n",
       "      <td>1</td>\n",
       "      <td>73</td>\n",
       "      <td>2016-01-01 20:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>10</td>\n",
       "      <td>26729</td>\n",
       "      <td>2016-01-01 20:18:43</td>\n",
       "      <td>0</td>\n",
       "      <td>74</td>\n",
       "      <td>2016-01-01 20:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>11</td>\n",
       "      <td>33853</td>\n",
       "      <td>2016-01-01 23:31:51</td>\n",
       "      <td>0</td>\n",
       "      <td>25</td>\n",
       "      <td>2016-01-01 23:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>12</td>\n",
       "      <td>33853</td>\n",
       "      <td>2016-01-01 23:36:46</td>\n",
       "      <td>0</td>\n",
       "      <td>26</td>\n",
       "      <td>2016-01-01 23:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>13</td>\n",
       "      <td>33853</td>\n",
       "      <td>2016-01-01 23:45:58</td>\n",
       "      <td>0</td>\n",
       "      <td>27</td>\n",
       "      <td>2016-01-01 23:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>14</td>\n",
       "      <td>17108</td>\n",
       "      <td>2016-01-02 02:35:42</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2016-01-02 02:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>15</td>\n",
       "      <td>30755</td>\n",
       "      <td>2016-01-02 04:40:40</td>\n",
       "      <td>0</td>\n",
       "      <td>15</td>\n",
       "      <td>2016-01-02 04:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>16</td>\n",
       "      <td>56584</td>\n",
       "      <td>2016-01-02 04:55:32</td>\n",
       "      <td>0</td>\n",
       "      <td>14</td>\n",
       "      <td>2016-01-02 04:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>17</td>\n",
       "      <td>37073</td>\n",
       "      <td>2016-01-02 06:37:00</td>\n",
       "      <td>0</td>\n",
       "      <td>78</td>\n",
       "      <td>2016-01-02 06:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>18</td>\n",
       "      <td>50166</td>\n",
       "      <td>2016-01-02 12:11:16</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>2016-01-02 12:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>19</td>\n",
       "      <td>56132</td>\n",
       "      <td>2016-01-02 12:59:55</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2016-01-02 12:00:00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    order_id  customer_id          order_date  is_canceled  \\\n",
       "0          0        52285 2016-01-01 00:09:51            0   \n",
       "1          1        58296 2016-01-01 14:36:14            0   \n",
       "2          2        30609 2016-01-01 14:47:46            0   \n",
       "3          3        30256 2016-01-01 15:06:36            0   \n",
       "4          4        52642 2016-01-01 16:27:37            0   \n",
       "5          5        56257 2016-01-01 17:14:02            0   \n",
       "6          6         9948 2016-01-01 18:45:36            0   \n",
       "7          7        39536 2016-01-01 19:17:55            0   \n",
       "8          8        26729 2016-01-01 19:47:50            1   \n",
       "9          9        26729 2016-01-01 20:01:01            1   \n",
       "10        10        26729 2016-01-01 20:18:43            0   \n",
       "11        11        33853 2016-01-01 23:31:51            0   \n",
       "12        12        33853 2016-01-01 23:36:46            0   \n",
       "13        13        33853 2016-01-01 23:45:58            0   \n",
       "14        14        17108 2016-01-02 02:35:42            0   \n",
       "15        15        30755 2016-01-02 04:40:40            0   \n",
       "16        16        56584 2016-01-02 04:55:32            0   \n",
       "17        17        37073 2016-01-02 06:37:00            0   \n",
       "18        18        50166 2016-01-02 12:11:16            0   \n",
       "19        19        56132 2016-01-02 12:59:55            0   \n",
       "\n",
       "    number_of_paid_orders_before     date_minus_time  \n",
       "0                              1 2016-01-01 00:00:00  \n",
       "1                              0 2016-01-01 14:00:00  \n",
       "2                              2 2016-01-01 14:00:00  \n",
       "3                              2 2016-01-01 15:00:00  \n",
       "4                              5 2016-01-01 16:00:00  \n",
       "5                              5 2016-01-01 17:00:00  \n",
       "6                            176 2016-01-01 18:00:00  \n",
       "7                             11 2016-01-01 19:00:00  \n",
       "8                             72 2016-01-01 19:00:00  \n",
       "9                             73 2016-01-01 20:00:00  \n",
       "10                            74 2016-01-01 20:00:00  \n",
       "11                            25 2016-01-01 23:00:00  \n",
       "12                            26 2016-01-01 23:00:00  \n",
       "13                            27 2016-01-01 23:00:00  \n",
       "14                             1 2016-01-02 02:00:00  \n",
       "15                            15 2016-01-02 04:00:00  \n",
       "16                            14 2016-01-02 04:00:00  \n",
       "17                            78 2016-01-02 06:00:00  \n",
       "18                             6 2016-01-02 12:00:00  \n",
       "19                             1 2016-01-02 12:00:00  "
      ]
     },
     "execution_count": 536,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head(20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "data['date_minus_time'] = data[\"order_date\"].apply( lambda df : \n",
    "                                datetime.datetime(year=df.year, month=df.month, day=df.day, hour=df.hour))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 382,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "date_minus_time\n",
       "1    1476\n",
       "2    2155\n",
       "3    2369\n",
       "4    3920\n",
       "5    3450\n",
       "6    2684\n",
       "Name: order_id, dtype: int64"
      ]
     },
     "execution_count": 382,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.groupby(data['date_minus_time'].dt.month)['order_id'].count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 383,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f050a2a48d0>"
      ]
     },
     "execution_count": 383,
     "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": [
    "data.groupby(data['date_minus_time'].dt.month)['order_id'].count().plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##  sales grow from 1 month with trend to continue"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 384,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "date_minus_time\n",
       "0      868\n",
       "1      987\n",
       "2     1045\n",
       "3      913\n",
       "4      808\n",
       "5      643\n",
       "6      410\n",
       "7      308\n",
       "8      279\n",
       "9      186\n",
       "10     213\n",
       "11     268\n",
       "12     311\n",
       "13     462\n",
       "14     595\n",
       "15     746\n",
       "16     744\n",
       "17     831\n",
       "18     872\n",
       "19     879\n",
       "20     896\n",
       "21     909\n",
       "22     941\n",
       "23     940\n",
       "Name: order_id, dtype: int64"
      ]
     },
     "execution_count": 384,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.groupby(data['date_minus_time'].dt.hour)['order_id'].count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 385,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f0509b28f98>"
      ]
     },
     "execution_count": 385,
     "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": [
    "data.groupby(data['date_minus_time'].dt.hour)['order_id'].count().plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### we see several peacks and deep drop of the seals "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 386,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "date_minus_time\n",
       "1      264\n",
       "2      295\n",
       "3      440\n",
       "4      421\n",
       "5      499\n",
       "6      472\n",
       "7      545\n",
       "8      557\n",
       "9      486\n",
       "10     526\n",
       "11     534\n",
       "12     533\n",
       "13     608\n",
       "14     776\n",
       "15     889\n",
       "16    1096\n",
       "17    1076\n",
       "18     952\n",
       "19     803\n",
       "20     720\n",
       "21     629\n",
       "22     694\n",
       "23     725\n",
       "24     582\n",
       "25     546\n",
       "26     330\n",
       "53      56\n",
       "Name: order_id, dtype: int64"
      ]
     },
     "execution_count": 386,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.groupby(data['date_minus_time'].dt.week)['order_id'].count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 387,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f05091fe748>"
      ]
     },
     "execution_count": 387,
     "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": [
    "data.groupby(data['date_minus_time'].dt.week)['order_id'].count().plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# After peak at fist week and drop mean sales relatively stable"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 388,
   "metadata": {},
   "outputs": [],
   "source": [
    "new = data[['date_minus_time','number_of_paid_orders_before']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 389,
   "metadata": {},
   "outputs": [],
   "source": [
    "new = new.set_index('date_minus_time')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 390,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.DataFrame({'data': data['number_of_paid_orders_before']}, index=data['date_minus_time'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 396,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 0.99729145 -2.1034977   0.28513867 ... -3.2908083  -0.74589037\n",
      " -0.71650275]\n",
      "[1 0 2 ... 3 6 2]\n"
     ]
    }
   ],
   "source": [
    "from statsmodels.tsa.seasonal import seasonal_decompose\n",
    "\n",
    "result = seasonal_decompose(new.number_of_paid_orders_before.values, freq=180, model='additive')\n",
    "\n",
    "print(result.seasonal)\n",
    "print(result.observed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 395,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 1440x1440 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(20, 20))\n",
    "result.plot()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### There is strong trend in the begging of the sales in 6 month period time. Not observed anything in 30 days period\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###     2.1. calculate cancel_rate for the whole dataset. What about standard error, confidence intervalsm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 397,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.07101034010215522"
      ]
     },
     "execution_count": 397,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.is_canceled.mean() # mean cansel rate of whole data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 398,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "confidence interval of the mean is from 0.06703700271067727 to 0.07498367749363317 \n",
      " standart error 0.0020270973555418684\n"
     ]
    }
   ],
   "source": [
    "se_ = np.std (data.is_canceled)/ np.sqrt(len(data))\n",
    "\n",
    "mean_ = data.is_canceled.mean()\n",
    "\n",
    "t_score = st.t.ppf(1-0.05/2, len(data) - 1)\n",
    "\n",
    "lower_t = mean_ - t_score * se_\n",
    "upper_t = mean_ + t_score * se_\n",
    "\n",
    "print('confidence interval of the mean is from {} to {} \\n standart error {}'.format(lower_t, upper_t, se_))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 399,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(20, 10))\n",
    "sns.distplot(data.is_canceled)\n",
    "plt.axvline(se_, color='green', linestyle='--')\n",
    "plt.axvline(lower_t, color='red')\n",
    "plt.axvline(upper_t, color='black')\n",
    "plt.legend(['lower bound', 'upper bound'])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " # investigate cancel_rate vs number_of_paid_orders_before (number_of_paid_orders_before altering in [0:10] range) – any ideaswhy dependency looks so? What about standard error?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 400,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.08535606820461385"
      ]
     },
     "execution_count": 400,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_f = data[data['number_of_paid_orders_before'] < 10]\n",
    "data_f.is_canceled.mean() # cansel rates for customers that buy less than 10 times"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 401,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>order_id</th>\n",
       "      <th>customer_id</th>\n",
       "      <th>order_date</th>\n",
       "      <th>is_canceled</th>\n",
       "      <th>number_of_paid_orders_before</th>\n",
       "      <th>date_minus_time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>52285</td>\n",
       "      <td>2016-01-01 00:09:51</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2016-01-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>58296</td>\n",
       "      <td>2016-01-01 14:36:14</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2016-01-01 14:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>30609</td>\n",
       "      <td>2016-01-01 14:47:46</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2016-01-01 14:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>30256</td>\n",
       "      <td>2016-01-01 15:06:36</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2016-01-01 15:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>52642</td>\n",
       "      <td>2016-01-01 16:27:37</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>2016-01-01 16:00:00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0  order_id  customer_id          order_date  is_canceled  \\\n",
       "0           0         0        52285 2016-01-01 00:09:51            0   \n",
       "1           1         1        58296 2016-01-01 14:36:14            0   \n",
       "2           2         2        30609 2016-01-01 14:47:46            0   \n",
       "3           3         3        30256 2016-01-01 15:06:36            0   \n",
       "4           4         4        52642 2016-01-01 16:27:37            0   \n",
       "\n",
       "   number_of_paid_orders_before     date_minus_time  \n",
       "0                             1 2016-01-01 00:00:00  \n",
       "1                             0 2016-01-01 14:00:00  \n",
       "2                             2 2016-01-01 14:00:00  \n",
       "3                             2 2016-01-01 15:00:00  \n",
       "4                             5 2016-01-01 16:00:00  "
      ]
     },
     "execution_count": 401,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_f.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 413,
   "metadata": {},
   "outputs": [],
   "source": [
    "me = data_f.groupby('number_of_paid_orders_before')['is_canceled'].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 422,
   "metadata": {},
   "outputs": [],
   "source": [
    "mean_n = []\n",
    "count_n = []\n",
    "std_error = []\n",
    "mean_cr = np.mean(me)\n",
    "alpha = 0.05\n",
    "t_score = st.t.ppf(1-alpha/2, 9)\n",
    "\n",
    "lower = []\n",
    "upper = []\n",
    "\n",
    "\n",
    "for i in range(0, 10):\n",
    "    df_temp = data_f[data_f['number_of_paid_orders_before'] == i]\n",
    "    mean_n.append(df_temp.is_canceled.mean())\n",
    "    count_n.append(len(df_temp))\n",
    "    std_error.append(df_temp.is_canceled.std()/np.sqrt(len(df_temp)))\n",
    "\n",
    "low = np.array(mean_n) - t_score * np.array(std_error)\n",
    "upp = np.array(mean_n) + t_score * np.array(std_error)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 432,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PolyCollection at 0x7f050b0b4240>"
      ]
     },
     "execution_count": 432,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(8, 6), dpi=80, facecolor='w', edgecolor='r')    \n",
    "plt.plot(me)\n",
    "plt.title('Cansel rate with CI')\n",
    "plt.xlabel('Months')\n",
    "plt.ylabel('Cansel rate')\n",
    "plt.fill_between(range(0,10), low, upp, edgecolor='none', facecolor='b', alpha=0.2, interpolate=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Cancel rate in month period"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 239,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f050a085198>"
      ]
     },
     "execution_count": 239,
     "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": [
    "data.groupby(data['date_minus_time'].dt.month)['is_canceled'].mean().plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Investigate time between customer's placed orders– dT (how often does the customer place orders)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 257,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>order_id</th>\n",
       "      <th>customer_id</th>\n",
       "      <th>order_date</th>\n",
       "      <th>is_canceled</th>\n",
       "      <th>number_of_paid_orders_before</th>\n",
       "      <th>date_minus_time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>52285</td>\n",
       "      <td>2016-01-01 00:09:51</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2016-01-01 00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>58296</td>\n",
       "      <td>2016-01-01 14:36:14</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2016-01-01 14:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>30609</td>\n",
       "      <td>2016-01-01 14:47:46</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2016-01-01 14:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>30256</td>\n",
       "      <td>2016-01-01 15:06:36</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2016-01-01 15:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>52642</td>\n",
       "      <td>2016-01-01 16:27:37</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>2016-01-01 16:00:00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0  order_id  customer_id          order_date  is_canceled  \\\n",
       "0           0         0        52285 2016-01-01 00:09:51            0   \n",
       "1           1         1        58296 2016-01-01 14:36:14            0   \n",
       "2           2         2        30609 2016-01-01 14:47:46            0   \n",
       "3           3         3        30256 2016-01-01 15:06:36            0   \n",
       "4           4         4        52642 2016-01-01 16:27:37            0   \n",
       "\n",
       "   number_of_paid_orders_before     date_minus_time  \n",
       "0                             1 2016-01-01 00:00:00  \n",
       "1                             0 2016-01-01 14:00:00  \n",
       "2                             2 2016-01-01 14:00:00  \n",
       "3                             2 2016-01-01 15:00:00  \n",
       "4                             5 2016-01-01 16:00:00  "
      ]
     },
     "execution_count": 257,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 433,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "customer_id\n",
       "7939    30\n",
       "7978     1\n",
       "8057     1\n",
       "8088     1\n",
       "8105     3\n",
       "Name: order_id, dtype: int64"
      ]
     },
     "execution_count": 433,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.groupby('customer_id')['order_id'].count().head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 675,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df_cst = data[data.customer_id == 7939].copy()\n",
    "# df_cst['time_next'] = df_cst.order_date.diff().shift(-1)\n",
    "\n",
    "df = data.sort_values('order_date')\n",
    "\n",
    "df['time_next'] = df.groupby('customer_id').order_date.transform(lambda x: x.diff().shift(-1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 674,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/volodymyr/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.py:1: FutureWarning: Passing integers to fillna is deprecated, will raise a TypeError in a future version.  To retain the old behavior, pass pd.Timedelta(seconds=n) instead.\n",
      "  \"\"\"Entry point for launching an IPython kernel.\n"
     ]
    }
   ],
   "source": [
    "df.fillna(0, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 495,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df[df['time_next'] != '0 days 00:00:00']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 676,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "8.974661544090743"
      ]
     },
     "execution_count": 676,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.time_next.dt.days.mean() # mean of the day"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 677,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/volodymyr/anaconda3/lib/python3.7/site-packages/numpy/lib/histograms.py:824: RuntimeWarning: invalid value encountered in greater_equal\n",
      "  keep = (tmp_a >= first_edge)\n",
      "/home/volodymyr/anaconda3/lib/python3.7/site-packages/numpy/lib/histograms.py:825: RuntimeWarning: invalid value encountered in less_equal\n",
      "  keep &= (tmp_a <= last_edge)\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": [
    "plt.hist(df.time_next.dt.total_seconds() / 3600, bins=50)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 678,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, 'orders')"
      ]
     },
     "execution_count": 678,
     "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.hist(df.time_next.dt.days)\n",
    "plt.title('Number of days between orders of customers')\n",
    "plt.xlabel('orders')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 679,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.1           0 days 00:13:08\n",
       "0.2    0 days 00:47:39.600000\n",
       "0.3    0 days 18:46:29.599999\n",
       "0.4           1 days 23:29:49\n",
       "0.5           3 days 19:13:58\n",
       "0.6    6 days 00:01:54.799999\n",
       "0.7    8 days 09:22:18.599999\n",
       "0.9   25 days 01:48:24.099999\n",
       "Name: time_next, dtype: timedelta64[ns]"
      ]
     },
     "execution_count": 679,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.time_next.quantile([0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.9])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 680,
   "metadata": {},
   "outputs": [],
   "source": [
    "del df['date_minus_time']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 681,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.sort_values('time_next', inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 682,
   "metadata": {},
   "outputs": [],
   "source": [
    "customers_days = df.groupby('customer_id')['time_next'].sum().sort_values(ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 683,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.1           0 days 00:00:00\n",
       "0.2           0 days 00:00:00\n",
       "0.3           0 days 00:00:00\n",
       "0.4           0 days 00:00:00\n",
       "0.5           0 days 00:00:00\n",
       "0.6    0 days 17:37:01.999999\n",
       "0.7   10 days 21:08:32.300000\n",
       "0.9   80 days 05:15:55.800000\n",
       "Name: time_next, dtype: timedelta64[ns]"
      ]
     },
     "execution_count": 683,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "customers_days.quantile([0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.9])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.2. compare retention time period for FCO (1st customer’s order) vs not FCO (2nd customer’s order and more)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 684,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "time_next  customer_id\n",
       "00:05:34   51721          2\n",
       "00:01:48   22814          2\n",
       "00:03:07   22814          2\n",
       "00:06:02   52363          2\n",
       "00:06:24   33853          2\n",
       "Name: customer_id, dtype: int64"
      ]
     },
     "execution_count": 684,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.groupby('time_next')['customer_id'].value_counts().sort_values(ascending=False).head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 685,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "datetime.timedelta(seconds=3600)"
      ]
     },
     "execution_count": 685,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import datetime\n",
    "dt_mult = datetime.timedelta(hours=1)\n",
    "dt_mult"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 686,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.14239441883642706"
      ]
     },
     "execution_count": 686,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(dt_mult > df.time_next).mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 687,
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy.stats import ttest_ind"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 688,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Ttest_indResult(statistic=-12.29627750876382, pvalue=2.8304097694236053e-34)"
      ]
     },
     "execution_count": 688,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data1 = df[df.number_of_paid_orders_before == 0].time_next.astype(int)\n",
    "data2 = df[df.number_of_paid_orders_before == 1].time_next.astype(int)\n",
    "\n",
    "ttest_ind(data1, data2, nan_policy='omit')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 689,
   "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>order_id</th>\n",
       "      <th>customer_id</th>\n",
       "      <th>order_date</th>\n",
       "      <th>is_canceled</th>\n",
       "      <th>number_of_paid_orders_before</th>\n",
       "      <th>time_next</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>852</th>\n",
       "      <td>852</td>\n",
       "      <td>35435</td>\n",
       "      <td>2016-01-21 15:05:29</td>\n",
       "      <td>0</td>\n",
       "      <td>13</td>\n",
       "      <td>00:00:04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>851</th>\n",
       "      <td>851</td>\n",
       "      <td>35435</td>\n",
       "      <td>2016-01-21 15:05:23</td>\n",
       "      <td>0</td>\n",
       "      <td>12</td>\n",
       "      <td>00:00:06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8348</th>\n",
       "      <td>8348</td>\n",
       "      <td>45004</td>\n",
       "      <td>2016-04-20 12:35:50</td>\n",
       "      <td>0</td>\n",
       "      <td>108</td>\n",
       "      <td>00:00:13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5883</th>\n",
       "      <td>5883</td>\n",
       "      <td>61188</td>\n",
       "      <td>2016-03-30 21:00:39</td>\n",
       "      <td>0</td>\n",
       "      <td>25</td>\n",
       "      <td>00:00:20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5884</th>\n",
       "      <td>5884</td>\n",
       "      <td>61188</td>\n",
       "      <td>2016-03-30 21:00:59</td>\n",
       "      <td>0</td>\n",
       "      <td>26</td>\n",
       "      <td>00:00:20</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      order_id  customer_id          order_date  is_canceled  \\\n",
       "852        852        35435 2016-01-21 15:05:29            0   \n",
       "851        851        35435 2016-01-21 15:05:23            0   \n",
       "8348      8348        45004 2016-04-20 12:35:50            0   \n",
       "5883      5883        61188 2016-03-30 21:00:39            0   \n",
       "5884      5884        61188 2016-03-30 21:00:59            0   \n",
       "\n",
       "      number_of_paid_orders_before time_next  \n",
       "852                             13  00:00:04  \n",
       "851                             12  00:00:06  \n",
       "8348                           108  00:00:13  \n",
       "5883                            25  00:00:20  \n",
       "5884                            26  00:00:20  "
      ]
     },
     "execution_count": 689,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 690,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Timedelta('9 days 20:18:00.540519')"
      ]
     },
     "execution_count": 690,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df.number_of_paid_orders_before == 0].time_next.mean() # People who buy first comes after 12 days"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 691,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Timedelta('11 days 03:20:47.345294')"
      ]
     },
     "execution_count": 691,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df.number_of_paid_orders_before <= 2].time_next.mean() #  After 11 days they buy second item"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3.3. investigate what is the fraction of ‘multi-orders’ for which the time between two sequential orders < dT_multi, and it is short enough to consider two orders to be placed simultaneously. What dT_multi would would you choose and why?    T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 692,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.set_index('order_date')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 693,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['dT_multi'] = df.time_next.transform(lambda x: x.diff().shift(-1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 694,
   "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>order_id</th>\n",
       "      <th>customer_id</th>\n",
       "      <th>is_canceled</th>\n",
       "      <th>number_of_paid_orders_before</th>\n",
       "      <th>time_next</th>\n",
       "      <th>dT_multi</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>order_date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2016-01-21 15:05:29</th>\n",
       "      <td>852</td>\n",
       "      <td>35435</td>\n",
       "      <td>0</td>\n",
       "      <td>13</td>\n",
       "      <td>00:00:04</td>\n",
       "      <td>00:00:02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-01-21 15:05:23</th>\n",
       "      <td>851</td>\n",
       "      <td>35435</td>\n",
       "      <td>0</td>\n",
       "      <td>12</td>\n",
       "      <td>00:00:06</td>\n",
       "      <td>00:00:07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-04-20 12:35:50</th>\n",
       "      <td>8348</td>\n",
       "      <td>45004</td>\n",
       "      <td>0</td>\n",
       "      <td>108</td>\n",
       "      <td>00:00:13</td>\n",
       "      <td>00:00:07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-03-30 21:00:39</th>\n",
       "      <td>5883</td>\n",
       "      <td>61188</td>\n",
       "      <td>0</td>\n",
       "      <td>25</td>\n",
       "      <td>00:00:20</td>\n",
       "      <td>00:00:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-03-30 21:00:59</th>\n",
       "      <td>5884</td>\n",
       "      <td>61188</td>\n",
       "      <td>0</td>\n",
       "      <td>26</td>\n",
       "      <td>00:00:20</td>\n",
       "      <td>00:00:02</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     order_id  customer_id  is_canceled  \\\n",
       "order_date                                                \n",
       "2016-01-21 15:05:29       852        35435            0   \n",
       "2016-01-21 15:05:23       851        35435            0   \n",
       "2016-04-20 12:35:50      8348        45004            0   \n",
       "2016-03-30 21:00:39      5883        61188            0   \n",
       "2016-03-30 21:00:59      5884        61188            0   \n",
       "\n",
       "                     number_of_paid_orders_before time_next dT_multi  \n",
       "order_date                                                            \n",
       "2016-01-21 15:05:29                            13  00:00:04 00:00:02  \n",
       "2016-01-21 15:05:23                            12  00:00:06 00:00:07  \n",
       "2016-04-20 12:35:50                           108  00:00:13 00:00:07  \n",
       "2016-03-30 21:00:39                            25  00:00:20 00:00:00  \n",
       "2016-03-30 21:00:59                            26  00:00:20 00:00:02  "
      ]
     },
     "execution_count": 694,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 695,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, '# hours')"
      ]
     },
     "execution_count": 695,
     "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": [
    "df['dT_multi'].quantile([0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.9]).plot()\n",
    "plt.xlabel('quantiles')\n",
    "plt.ylabel('# hours')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### We can count 0.2 sec as time after we may consider a not simultanious order"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "4. Investigate customer retention_rate: % of customers who return after completing an order"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 698,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.2924885274937606"
      ]
     },
     "execution_count": 698,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(df[df.number_of_paid_orders_before == 0]) / len(df[df.number_of_paid_orders_before > 0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 699,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.reset_index(inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 700,
   "metadata": {},
   "outputs": [],
   "source": [
    "dt_loss = datetime.timedelta(days=30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 701,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['returned_after'] = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 702,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.loc[df.time_next < dt_loss, 'returned_after'] = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 705,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6282546405880155"
      ]
     },
     "execution_count": 705,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.returned_after.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 713,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.047805794663911653\n"
     ]
    }
   ],
   "source": [
    "mask = (df.is_canceled == 1)\n",
    "\n",
    "_, pvalue = ttest_ind(df[mask].returned_after, df[~mask].returned_after, nan_policy='omit')\n",
    "\n",
    "pvalue /=2\n",
    "\n",
    "print(pvalue) # less than 0.05 - so it is stat sagnificant"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 714,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.6052631578947368, 0.6300120691967279)"
      ]
     },
     "execution_count": 714,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[mask].returned_after.mean(), df[~mask].returned_after.mean() # мі проверяем "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 715,
   "metadata": {},
   "outputs": [],
   "source": [
    "mask_1 = (df.number_of_paid_orders_before <= 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df[mask_1 & mask]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 720,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.13157894736842105"
      ]
     },
     "execution_count": 720,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(df[mask].time_next < dt_mult).mean() # проверяем причины возврата товара и возможной новой покупки "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 717,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.14322113450449242"
      ]
     },
     "execution_count": 717,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(df[~mask].time_next < dt_mult).mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 721,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Timedelta('1 days 12:37:07')"
      ]
     },
     "execution_count": 721,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[mask].time_next.median()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 722,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Timedelta('3 days 23:29:22.500000')"
      ]
     },
     "execution_count": 722,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[~mask].time_next.median()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}