final.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 226,
   "id": "subtle-public",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.decomposition import TruncatedSVD\n",
    "from scipy.sparse.linalg import svds\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 227,
   "id": "informal-automation",
   "metadata": {},
   "outputs": [],
   "source": [
    "import cx_Oracle \n",
    "from pandas import Series, DataFrame\n",
    "import pandas as pd\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 228,
   "id": "cleared-university",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   lesson_seq                        lesson_title\n",
      "0           6       어제 폭식했니? 급찐급빠로 이 언니가 모조리 혼내줄게\n",
      "1           4                         2강 박상빈의 이해 \n",
      "2           8                  7월 2일 테스트 강의 등록입니다\n",
      "3          10         빈디젤의 workout are U ready???\n",
      "4          11                    GYM 종국와 빈디젤의 콜라보\n",
      "5          12                      강아지와 함께하는 맨몸운동\n",
      "6          14                         내가바로JAYPARK\n",
      "7          15                    정신똑바로 차리세요 BBAM!\n",
      "8          16                    맨몸운동의 정석 : 코어의 힘\n",
      "9           2                               ddddd\n",
      "10         17                      사랑이와 필라테스의 세계로\n",
      "11         18                       누워서 먹기만한 그대에게\n",
      "12         19                            헬스는 백년운동\n",
      "13         20                             asdfsdf\n",
      "14         21  스트레스와 다이어트를 한번에 해결하는 all-in-one 운동\n",
      "15         23                       New 뉴 근육운동가이드\n"
     ]
    }
   ],
   "source": [
    "connection = cx_Oracle.connect('profit/java@192.168.41.6:1521/xe')\n",
    "cursor = connection.cursor()\n",
    "cursor.execute(\"\"\"\n",
    "        select a.lesson_seq, a.lesson_title\n",
    "        from lesson a, lesson_category b\n",
    "        where a.lesson_category_seq = b.lesson_category_seq(+)\n",
    "        and lesson_private_flag = 'N'\n",
    "        \"\"\")\n",
    "\n",
    "result = cursor.fetchall()\n",
    "lesson_data = pd.DataFrame(result)\n",
    "lesson_data.rename(columns = {0 : 'lesson_seq'}, inplace = True)\n",
    "lesson_data.rename(columns = {1 : 'lesson_title'}, inplace = True)\n",
    "print(lesson_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 229,
   "id": "committed-colony",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "   user_id lesson_seq  rating\n",
      "0   S00001          3     5.0\n",
      "1   S00001         14     5.0\n",
      "2   S00001         16     5.0\n",
      "3   S00001          8     5.0\n",
      "4   S00002         21     5.0\n",
      "..     ...        ...     ...\n",
      "64    a003         21     3.5\n",
      "65    a003         23     3.5\n",
      "66    a004         20     3.5\n",
      "67    a004         21     3.5\n",
      "68    a004         23     3.5\n",
      "\n",
      "[69 rows x 3 columns]\n"
     ]
    }
   ],
   "source": [
    "cursor.execute(\"\"\"\n",
    "        select in_user_id as user_id, lesson_seq, 5 as rating\n",
    "        from buy_lesson\n",
    "        where lesson_refund_flag = 'N'\n",
    "        \n",
    "        union all\n",
    "        \n",
    "        select user_id, lesson_seq, 2.5 as rating\n",
    "        from\n",
    "        (\n",
    "        select in_user_id as user_id, lesson_seq\n",
    "        from bookgood\n",
    "        where bookgood_flag = 'Y'\n",
    "        and bookgood_gubun = 'B'\n",
    "        and lesson_seq is not null\n",
    "        MINUS\n",
    "        select in_user_id as user_id, lesson_seq\n",
    "        from buy_lesson\n",
    "        where lesson_refund_flag = 'N'\n",
    "        MINUS\n",
    "        select in_user_id as user_id, lesson_seq\n",
    "        from bookgood\n",
    "        where bookgood_flag='Y'\n",
    "        and bookgood_gubun = 'G'\n",
    "        and lesson_seq is not null\n",
    "        )\n",
    "        \n",
    "        union all\n",
    "        \n",
    "        select user_id, lesson_seq, 2.5 as rating\n",
    "        from\n",
    "        (\n",
    "        select in_user_id as user_id, lesson_seq\n",
    "        from bookgood\n",
    "        where bookgood_flag='Y'\n",
    "        and bookgood_gubun = 'G'\n",
    "        and lesson_seq is not null\n",
    "        MINUS\n",
    "        select in_user_id as user_id, lesson_seq\n",
    "        from buy_lesson\n",
    "        where lesson_refund_flag = 'N'\n",
    "        MINUS\n",
    "        select in_user_id as user_id, lesson_seq\n",
    "        from bookgood\n",
    "        where bookgood_flag = 'Y'\n",
    "        and bookgood_gubun = 'B'\n",
    "        and lesson_seq is not null\n",
    "        )\n",
    "        \n",
    "        union all\n",
    "        \n",
    "        select user_id, lesson_seq, 3.5 as rating\n",
    "        from\n",
    "        (\n",
    "        select in_user_id as user_id, lesson_seq\n",
    "        from bookgood\n",
    "        where bookgood_flag='Y'\n",
    "        and bookgood_gubun = 'G'\n",
    "        and lesson_seq is not null\n",
    "        INTERSECT\n",
    "        select in_user_id as user_id, lesson_seq\n",
    "        from bookgood\n",
    "        where bookgood_flag = 'Y'\n",
    "        and bookgood_gubun = 'B'\n",
    "        and lesson_seq is not null\n",
    "        MINUS\n",
    "        select in_user_id as user_id, lesson_seq\n",
    "        from buy_lesson\n",
    "        where lesson_refund_flag = 'N'\n",
    "        )\n",
    "        \"\"\")\n",
    "\n",
    "result1 = cursor.fetchall()\n",
    "rating_data = pd.DataFrame(result1)\n",
    "rating_data.rename(columns = {0 : 'user_id'}, inplace = True)\n",
    "rating_data.rename(columns = {1 : 'lesson_seq'}, inplace = True)\n",
    "rating_data.rename(columns = {2 : 'rating'}, inplace = True)\n",
    "print(rating_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 230,
   "id": "hydraulic-league",
   "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>user_id</th>\n",
       "      <th>lesson_seq</th>\n",
       "      <th>rating</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>S00001</td>\n",
       "      <td>3</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>S00001</td>\n",
       "      <td>14</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>S00001</td>\n",
       "      <td>16</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>S00001</td>\n",
       "      <td>8</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>S00002</td>\n",
       "      <td>21</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  user_id lesson_seq  rating\n",
       "0  S00001          3     5.0\n",
       "1  S00001         14     5.0\n",
       "2  S00001         16     5.0\n",
       "3  S00001          8     5.0\n",
       "4  S00002         21     5.0"
      ]
     },
     "execution_count": 230,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rating_data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 231,
   "id": "convinced-adjustment",
   "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>lesson_seq</th>\n",
       "      <th>lesson_title</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>6</td>\n",
       "      <td>어제 폭식했니? 급찐급빠로 이 언니가 모조리 혼내줄게</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4</td>\n",
       "      <td>2강 박상빈의 이해</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>8</td>\n",
       "      <td>7월 2일 테스트 강의 등록입니다</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>10</td>\n",
       "      <td>빈디젤의 workout are U ready???</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>11</td>\n",
       "      <td>GYM 종국와 빈디젤의 콜라보</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  lesson_seq                   lesson_title\n",
       "0          6  어제 폭식했니? 급찐급빠로 이 언니가 모조리 혼내줄게\n",
       "1          4                    2강 박상빈의 이해 \n",
       "2          8             7월 2일 테스트 강의 등록입니다\n",
       "3         10    빈디젤의 workout are U ready???\n",
       "4         11               GYM 종국와 빈디젤의 콜라보"
      ]
     },
     "execution_count": 231,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lesson_data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 232,
   "id": "moral-march",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(16, 2)\n",
      "(69, 3)\n"
     ]
    }
   ],
   "source": [
    "print(lesson_data.shape)\n",
    "print(rating_data.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 233,
   "id": "plastic-compatibility",
   "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>user_id</th>\n",
       "      <th>lesson_seq</th>\n",
       "      <th>rating</th>\n",
       "      <th>lesson_title</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>S00001</td>\n",
       "      <td>14</td>\n",
       "      <td>5.0</td>\n",
       "      <td>내가바로JAYPARK</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>a002</td>\n",
       "      <td>14</td>\n",
       "      <td>3.5</td>\n",
       "      <td>내가바로JAYPARK</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>a003</td>\n",
       "      <td>14</td>\n",
       "      <td>3.5</td>\n",
       "      <td>내가바로JAYPARK</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>S00001</td>\n",
       "      <td>16</td>\n",
       "      <td>5.0</td>\n",
       "      <td>맨몸운동의 정석 : 코어의 힘</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>a002</td>\n",
       "      <td>16</td>\n",
       "      <td>3.5</td>\n",
       "      <td>맨몸운동의 정석 : 코어의 힘</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  user_id lesson_seq  rating      lesson_title\n",
       "0  S00001         14     5.0       내가바로JAYPARK\n",
       "1    a002         14     3.5       내가바로JAYPARK\n",
       "2    a003         14     3.5       내가바로JAYPARK\n",
       "3  S00001         16     5.0  맨몸운동의 정석 : 코어의 힘\n",
       "4    a002         16     3.5  맨몸운동의 정석 : 코어의 힘"
      ]
     },
     "execution_count": 233,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_lesson_data = pd.merge(rating_data, lesson_data, on = 'lesson_seq')\n",
    "user_lesson_data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "id": "sexual-costa",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(65, 4)"
      ]
     },
     "execution_count": 234,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_lesson_data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 235,
   "id": "premier-creation",
   "metadata": {},
   "outputs": [],
   "source": [
    "user_lesson_rating = user_lesson_data.pivot_table('rating', index = 'user_id', columns='lesson_seq').fillna(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 236,
   "id": "marine-killer",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(16, 12)"
      ]
     },
     "execution_count": 236,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_lesson_rating.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 237,
   "id": "manual-opinion",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(16, 12)"
      ]
     },
     "execution_count": 237,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_lesson_rating.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 238,
   "id": "bacterial-webster",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(12, 16)"
      ]
     },
     "execution_count": 238,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lesson_user_rating = user_lesson_rating.values.T\n",
    "lesson_user_rating.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 239,
   "id": "smoking-paste",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "numpy.ndarray"
      ]
     },
     "execution_count": 239,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "type(lesson_user_rating)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 240,
   "id": "homeless-brief",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(12, 12)"
      ]
     },
     "execution_count": 240,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "SVD = TruncatedSVD(n_components=12)\n",
    "matrix = SVD.fit_transform(lesson_user_rating)\n",
    "matrix.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 241,
   "id": "economic-sequence",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 7.48074308e-01, -1.71415496e+00,  8.72757660e-01,  8.05845185e-01,\n",
       "        5.20188945e-01, -5.72487879e-01, -5.21182862e-01, -4.50851242e-01,\n",
       "       -4.89235235e-01, -1.68454068e-01,  1.07132418e-17, -8.52818612e-32])"
      ]
     },
     "execution_count": 241,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "matrix[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 242,
   "id": "infrared-editing",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(12, 12)"
      ]
     },
     "execution_count": 242,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corr = np.corrcoef(matrix)\n",
    "corr.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 243,
   "id": "narrow-format",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(12, 12)"
      ]
     },
     "execution_count": 243,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corr2 = corr[:200, :200]\n",
    "corr2.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 244,
   "id": "clean-addiction",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 244,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(16, 10))\n",
    "sns.heatmap(corr2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 245,
   "id": "sound-maine",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['11', '14', '15', '16', '17', '18', '19', '2', '20', '21', '23', '8']"
      ]
     },
     "execution_count": 245,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lesson_seq = user_lesson_rating.columns\n",
    "lesson_seq_list = list(lesson_seq)\n",
    "lesson_seq_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 246,
   "id": "hungarian-diagram",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[['16', '14', '20', '2', '18'], ['16', '20', '8', '11', '18'], ['18', '19', '17', '14', '21'], ['14', '8', '11', '20', '18'], ['18', '21', '19', '23', '2'], ['20', '23', '21', '17', '19'], ['18', '17', '20', '23', '15'], ['17', '11', '8', '21', '20'], ['18', '21', '14', '16', '8'], ['20', '18', '23', '17', '14'], ['18', '21', '17', '19', '14'], ['16', '14', '20', '2', '18']]\n"
     ]
    }
   ],
   "source": [
    "array = []\n",
    "for i in range(len(lesson_seq_list)):\n",
    "    coffey_hands = i\n",
    "    corr_coffey_hands  = corr[coffey_hands]\n",
    "    result_list = []\n",
    "    per = 0.9\n",
    "    while(per >= 0.1):\n",
    "        temp1 = list(lesson_seq[(corr_coffey_hands >= per)])[:50]\n",
    "        temp2 = list(lesson_seq[(corr_coffey_hands >= per-0.1)])[:50]\n",
    "        temp3 = list(set(temp2) - set(temp1))\n",
    "        cnt = 0\n",
    "        for j in temp3:\n",
    "            if len(result_list) < 5:\n",
    "                result_list.append(j)\n",
    "        per = per - 0.1\n",
    "        cnt = cnt + 1\n",
    "    array.append(result_list)\n",
    "print(array)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 247,
   "id": "governing-newton",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1"
      ]
     },
     "execution_count": 247,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cursor.execute(\n",
    "        \"\"\"\n",
    "        select nvl(max(recommand_seq)+1, 1)\n",
    "        from recommand\n",
    "        \"\"\")\n",
    "select = cursor.fetchall()\n",
    "select_value = pd.DataFrame(select)\n",
    "select_val = int(select_value.iloc[0][0])\n",
    "select_val"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 248,
   "id": "becoming-concentrate",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "for i in range(len(array)):\n",
    "    for j in range(len(array[i])):\n",
    "        cursor.execute(\n",
    "        \"\"\"\n",
    "        insert into recommand(recommand_seq, recommand_detail_seq,\n",
    "        lesson_seq, in_date)\n",
    "        values\n",
    "        (\n",
    "        :rSeq, :rdSeq, :lessonSeq, sysdate)\n",
    "        \"\"\"\n",
    "        ,rSeq = select_val, rdSeq = array[i][j],lessonSeq = lesson_seq_list[i])\n",
    "        connection.commit()\n",
    "        select_val = select_val + 1\n",
    "        connection.commit()\n"
   ]
  }
 ],
 "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}