data_preprocessing.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "dc05285c",
   "metadata": {},
   "source": [
    "# Recipe Recommendation System: Data Preparation & Initial EDA"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f6abc337",
   "metadata": {},
   "source": [
    "Author: Kelly Li\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8753194c",
   "metadata": {},
   "source": [
    "## Table of contents:\n",
    "* [1 Introduction](#one)\n",
    "* [2 Datasets](#two)\n",
    "    * [2.1 Data Sources](#twoone)\n",
    "    * [2.2 Data Import & Summary](#twotwo)\n",
    "* [3 Data Cleaning](#three)\n",
    "    * [3.1 Data Types](#threeone) \n",
    "    * [3.2 Missing Data](#threetwo)\n",
    "    * [3.3 Duplicate Data](#threethree)\n",
    "    * [3.4 Outliers](#threefour)\n",
    "    * [3.5 Zero Ratings](#threefive)\n",
    "* [4 Data Preprocessing](#four)\n",
    "    * [4.1 Feature Engineering](#fourone)\n",
    "        * [4.1.1 Nutritional Values](#fouroneone)\n",
    "        * [4.1.2 Time Components](#fouronetwo)\n",
    "        * [4.1.3 Dietary Preference](#fouronethree)\n",
    "    * [4.2 Aggregation](#fourtwo)\n",
    "        * [4.2.1 Ratings](#fourtwoone)\n",
    "        * [4.2.2 Merge Datasets](#fourtwotwo)\n",
    "* [5 Conclusion](#five)\n",
    "\n",
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7b43e23a",
   "metadata": {},
   "source": [
    "## 1 Introduction <a class=\"anchor\" id=\"one\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1f2c33d4",
   "metadata": {},
   "source": [
    "Cooking enthusiasts often face challenges in finding personalized and diverse recipes that align with their tastes, dietar restrictions, and ingredient preferences. The existing search process is time-consuming, lacks inspiration, and fails to cater to specific dietary needs. To address these challenges, we have developed a recipe recommendation system that aids in recipe discovery, focusing on providing inspiration and catering to specific dietary needs. By leveraging advanced algorithms, it curates personalized recipe suggestions that ignite culinary creativity while considering unique tasts and dietary requirements. Wecome a world a culinary inspiration with"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2871e342",
   "metadata": {},
   "source": [
    "## 2 Datasets <a class=\"anchor\" id=\"two\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "98e19652",
   "metadata": {},
   "source": [
    "The food.com dataset, curated by Bodhisattwa Prasad Majumder, Shuyang Li, Jianmo Ni, and Julian McAuley, provides a wealth of information for our exploration study. With over 180,000 recipes and 700,000 recipe reviews, this dataset captures 18 years of user interactions and uploads on Food.com."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9ee7db14",
   "metadata": {},
   "source": [
    "### 2.1 Data Sources <a class=\"anchor\" id=\"twoone\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cee93547",
   "metadata": {},
   "source": [
    "Here's the data dictionary for the columns in the raw recipe dataset:\n",
    "\n",
    "| Column Name   | Description                                               |\n",
    "| ------------- | --------------------------------------------------------- |\n",
    "| name          | The name of the recipe.                                   |\n",
    "| id            | The unique identifier of the recipe.                       |\n",
    "| minutes       | The total cooking and preparation time in minutes.         |\n",
    "| contributor_id| The unique identifier of the user who submitted the recipe|\n",
    "| submitted     | The date when the recipe was submitted.                    |\n",
    "| tags          | Tags or labels associated with the recipe (e.g., vegetarian, vegan, dessert). |\n",
    "| nutrition     | Nutritional information (i.e. calories (#), total fat (PDV*), sugar (PDV*) , sodium (PDV*) , protein (PDV*) , saturated fat (PDV*), carbohydrates (PDV*)). |\n",
    "| n_steps       | The total number of steps or instructions in the recipe.   |\n",
    "| steps         | The step-by-step instructions for preparing the recipe.    |\n",
    "| description   | A brief description or summary of the recipe.              |\n",
    "| ingredients   | The list of ingredients required for the recipe.           |\n",
    "| n_ingredients | The total number of ingredients used in the recipe.        |\n",
    "\n",
    "*PDV = % daily value"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "07dd9dc6",
   "metadata": {},
   "source": [
    "Here's the data dictionary for the columns in the raw user interactions dataset:\n",
    "\n",
    "| Column Name | Description                                          |\n",
    "| ----------- | ---------------------------------------------------- |\n",
    "| user_id     | The unique identifier of the user.                    |\n",
    "| recipe_id   | The unique identifier of the recipe associated with the user's interaction. |\n",
    "| date        | The date when the user's interaction took place.      |\n",
    "| rating      | The rating given by the user for the recipe.          |\n",
    "| review      | The review or feedback provided by the user for the recipe. |"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f685213c",
   "metadata": {},
   "source": [
    "### 2.2 Data Import & Summary <a class=\"anchor\" id=\"twotwo\"></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "ea442b77",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import libraries\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "5a92abe3",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Read the raw datasets\n",
    "raw_recipes_df = pd.read_csv('RAW_recipes.csv')\n",
    "raw_users_df = pd.read_csv('RAW_interactions.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "62f39081",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The shape of the recipe dataset is: (231637, 12)\n",
      "The shape of the user interactions dataset is: (1132367, 5)\n"
     ]
    }
   ],
   "source": [
    "# Shape of the data\n",
    "print('The shape of the recipe dataset is:', raw_recipes_df.shape)\n",
    "print('The shape of the user interactions dataset is:', raw_users_df.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "cdf3805c",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>id</th>\n",
       "      <th>minutes</th>\n",
       "      <th>contributor_id</th>\n",
       "      <th>submitted</th>\n",
       "      <th>tags</th>\n",
       "      <th>nutrition</th>\n",
       "      <th>n_steps</th>\n",
       "      <th>steps</th>\n",
       "      <th>description</th>\n",
       "      <th>ingredients</th>\n",
       "      <th>n_ingredients</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>arriba   baked winter squash mexican style</td>\n",
       "      <td>137739</td>\n",
       "      <td>55</td>\n",
       "      <td>47892</td>\n",
       "      <td>2005-09-16</td>\n",
       "      <td>['60-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>[51.5, 0.0, 13.0, 0.0, 2.0, 0.0, 4.0]</td>\n",
       "      <td>11</td>\n",
       "      <td>['make a choice and proceed with recipe', 'dep...</td>\n",
       "      <td>autumn is my favorite time of year to cook! th...</td>\n",
       "      <td>['winter squash', 'mexican seasoning', 'mixed ...</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>a bit different  breakfast pizza</td>\n",
       "      <td>31490</td>\n",
       "      <td>30</td>\n",
       "      <td>26278</td>\n",
       "      <td>2002-06-17</td>\n",
       "      <td>['30-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>[173.4, 18.0, 0.0, 17.0, 22.0, 35.0, 1.0]</td>\n",
       "      <td>9</td>\n",
       "      <td>['preheat oven to 425 degrees f', 'press dough...</td>\n",
       "      <td>this recipe calls for the crust to be prebaked...</td>\n",
       "      <td>['prepared pizza crust', 'sausage patty', 'egg...</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>all in the kitchen  chili</td>\n",
       "      <td>112140</td>\n",
       "      <td>130</td>\n",
       "      <td>196586</td>\n",
       "      <td>2005-02-25</td>\n",
       "      <td>['time-to-make', 'course', 'preparation', 'mai...</td>\n",
       "      <td>[269.8, 22.0, 32.0, 48.0, 39.0, 27.0, 5.0]</td>\n",
       "      <td>6</td>\n",
       "      <td>['brown ground beef in large pot', 'add choppe...</td>\n",
       "      <td>this modified version of 'mom's' chili was a h...</td>\n",
       "      <td>['ground beef', 'yellow onions', 'diced tomato...</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>alouette  potatoes</td>\n",
       "      <td>59389</td>\n",
       "      <td>45</td>\n",
       "      <td>68585</td>\n",
       "      <td>2003-04-14</td>\n",
       "      <td>['60-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>[368.1, 17.0, 10.0, 2.0, 14.0, 8.0, 20.0]</td>\n",
       "      <td>11</td>\n",
       "      <td>['place potatoes in a large pot of lightly sal...</td>\n",
       "      <td>this is a super easy, great tasting, make ahea...</td>\n",
       "      <td>['spreadable cheese with garlic and herbs', 'n...</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>amish  tomato ketchup  for canning</td>\n",
       "      <td>44061</td>\n",
       "      <td>190</td>\n",
       "      <td>41706</td>\n",
       "      <td>2002-10-25</td>\n",
       "      <td>['weeknight', 'time-to-make', 'course', 'main-...</td>\n",
       "      <td>[352.9, 1.0, 337.0, 23.0, 3.0, 0.0, 28.0]</td>\n",
       "      <td>5</td>\n",
       "      <td>['mix all ingredients&amp; boil for 2 1 / 2 hours ...</td>\n",
       "      <td>my dh's amish mother raised him on this recipe...</td>\n",
       "      <td>['tomato juice', 'apple cider vinegar', 'sugar...</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         name      id  minutes  \\\n",
       "0  arriba   baked winter squash mexican style  137739       55   \n",
       "1            a bit different  breakfast pizza   31490       30   \n",
       "2                   all in the kitchen  chili  112140      130   \n",
       "3                          alouette  potatoes   59389       45   \n",
       "4          amish  tomato ketchup  for canning   44061      190   \n",
       "\n",
       "   contributor_id   submitted  \\\n",
       "0           47892  2005-09-16   \n",
       "1           26278  2002-06-17   \n",
       "2          196586  2005-02-25   \n",
       "3           68585  2003-04-14   \n",
       "4           41706  2002-10-25   \n",
       "\n",
       "                                                tags  \\\n",
       "0  ['60-minutes-or-less', 'time-to-make', 'course...   \n",
       "1  ['30-minutes-or-less', 'time-to-make', 'course...   \n",
       "2  ['time-to-make', 'course', 'preparation', 'mai...   \n",
       "3  ['60-minutes-or-less', 'time-to-make', 'course...   \n",
       "4  ['weeknight', 'time-to-make', 'course', 'main-...   \n",
       "\n",
       "                                    nutrition  n_steps  \\\n",
       "0       [51.5, 0.0, 13.0, 0.0, 2.0, 0.0, 4.0]       11   \n",
       "1   [173.4, 18.0, 0.0, 17.0, 22.0, 35.0, 1.0]        9   \n",
       "2  [269.8, 22.0, 32.0, 48.0, 39.0, 27.0, 5.0]        6   \n",
       "3   [368.1, 17.0, 10.0, 2.0, 14.0, 8.0, 20.0]       11   \n",
       "4   [352.9, 1.0, 337.0, 23.0, 3.0, 0.0, 28.0]        5   \n",
       "\n",
       "                                               steps  \\\n",
       "0  ['make a choice and proceed with recipe', 'dep...   \n",
       "1  ['preheat oven to 425 degrees f', 'press dough...   \n",
       "2  ['brown ground beef in large pot', 'add choppe...   \n",
       "3  ['place potatoes in a large pot of lightly sal...   \n",
       "4  ['mix all ingredients& boil for 2 1 / 2 hours ...   \n",
       "\n",
       "                                         description  \\\n",
       "0  autumn is my favorite time of year to cook! th...   \n",
       "1  this recipe calls for the crust to be prebaked...   \n",
       "2  this modified version of 'mom's' chili was a h...   \n",
       "3  this is a super easy, great tasting, make ahea...   \n",
       "4  my dh's amish mother raised him on this recipe...   \n",
       "\n",
       "                                         ingredients  n_ingredients  \n",
       "0  ['winter squash', 'mexican seasoning', 'mixed ...              7  \n",
       "1  ['prepared pizza crust', 'sausage patty', 'egg...              6  \n",
       "2  ['ground beef', 'yellow onions', 'diced tomato...             13  \n",
       "3  ['spreadable cheese with garlic and herbs', 'n...             11  \n",
       "4  ['tomato juice', 'apple cider vinegar', 'sugar...              8  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# View recipes dataset\n",
    "raw_recipes_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "9ccc0a7b",
   "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>recipe_id</th>\n",
       "      <th>date</th>\n",
       "      <th>rating</th>\n",
       "      <th>review</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>38094</td>\n",
       "      <td>40893</td>\n",
       "      <td>2003-02-17</td>\n",
       "      <td>4</td>\n",
       "      <td>Great with a salad. Cooked on top of stove for...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1293707</td>\n",
       "      <td>40893</td>\n",
       "      <td>2011-12-21</td>\n",
       "      <td>5</td>\n",
       "      <td>So simple, so delicious! Great for chilly fall...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>8937</td>\n",
       "      <td>44394</td>\n",
       "      <td>2002-12-01</td>\n",
       "      <td>4</td>\n",
       "      <td>This worked very well and is EASY.  I used not...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>126440</td>\n",
       "      <td>85009</td>\n",
       "      <td>2010-02-27</td>\n",
       "      <td>5</td>\n",
       "      <td>I made the Mexican topping and took it to bunk...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>57222</td>\n",
       "      <td>85009</td>\n",
       "      <td>2011-10-01</td>\n",
       "      <td>5</td>\n",
       "      <td>Made the cheddar bacon topping, adding a sprin...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   user_id  recipe_id        date  rating  \\\n",
       "0    38094      40893  2003-02-17       4   \n",
       "1  1293707      40893  2011-12-21       5   \n",
       "2     8937      44394  2002-12-01       4   \n",
       "3   126440      85009  2010-02-27       5   \n",
       "4    57222      85009  2011-10-01       5   \n",
       "\n",
       "                                              review  \n",
       "0  Great with a salad. Cooked on top of stove for...  \n",
       "1  So simple, so delicious! Great for chilly fall...  \n",
       "2  This worked very well and is EASY.  I used not...  \n",
       "3  I made the Mexican topping and took it to bunk...  \n",
       "4  Made the cheddar bacon topping, adding a sprin...  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# View user interactions dataset\n",
    "raw_users_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ad62a850",
   "metadata": {},
   "source": [
    "## 3 Data Cleaning <a class=\"anchor\" id=\"three\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4a7a5c75",
   "metadata": {},
   "source": [
    "### 3.1 Data Types <a class=\"anchor\" id=\"threeone\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f155061c",
   "metadata": {},
   "source": [
    "To gain a comprehensive understanding of the data types and identify any non-null values within the dataset, we will employ the `.info()` method."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "98f7cfb9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 231637 entries, 0 to 231636\n",
      "Data columns (total 12 columns):\n",
      " #   Column          Non-Null Count   Dtype \n",
      "---  ------          --------------   ----- \n",
      " 0   name            231636 non-null  object\n",
      " 1   id              231637 non-null  int64 \n",
      " 2   minutes         231637 non-null  int64 \n",
      " 3   contributor_id  231637 non-null  int64 \n",
      " 4   submitted       231637 non-null  object\n",
      " 5   tags            231637 non-null  object\n",
      " 6   nutrition       231637 non-null  object\n",
      " 7   n_steps         231637 non-null  int64 \n",
      " 8   steps           231637 non-null  object\n",
      " 9   description     226658 non-null  object\n",
      " 10  ingredients     231637 non-null  object\n",
      " 11  n_ingredients   231637 non-null  int64 \n",
      "dtypes: int64(5), object(7)\n",
      "memory usage: 21.2+ MB\n"
     ]
    }
   ],
   "source": [
    "# Info about the recipe dataset\n",
    "raw_recipes_df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e642785a",
   "metadata": {},
   "source": [
    "The 'id' and 'contributor_id' columns are currently of `integer` data type, while the 'submitted' column is of `object` data type. To ensure consistency and facilitate data manipulation, we will convert the 'id' and 'contributor_id' columns to `object` data type. Additionally, we will convert the 'submitted' column to `datetime` data type for more convenient date-based operations."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "30e00d51",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 231637 entries, 0 to 231636\n",
      "Data columns (total 12 columns):\n",
      " #   Column          Non-Null Count   Dtype         \n",
      "---  ------          --------------   -----         \n",
      " 0   name            231636 non-null  object        \n",
      " 1   id              231637 non-null  object        \n",
      " 2   minutes         231637 non-null  int64         \n",
      " 3   contributor_id  231637 non-null  object        \n",
      " 4   submitted       231637 non-null  datetime64[ns]\n",
      " 5   tags            231637 non-null  object        \n",
      " 6   nutrition       231637 non-null  object        \n",
      " 7   n_steps         231637 non-null  int64         \n",
      " 8   steps           231637 non-null  object        \n",
      " 9   description     226658 non-null  object        \n",
      " 10  ingredients     231637 non-null  object        \n",
      " 11  n_ingredients   231637 non-null  int64         \n",
      "dtypes: datetime64[ns](1), int64(3), object(8)\n",
      "memory usage: 21.2+ MB\n"
     ]
    }
   ],
   "source": [
    "# Convert 'id' and 'contributor_id' columns to object data type\n",
    "raw_recipes_df['id'] = raw_recipes_df['id'].astype(str)\n",
    "raw_recipes_df['contributor_id'] = raw_recipes_df['contributor_id'].astype(str)\n",
    "\n",
    "# Convert 'submitted' column to date data type\n",
    "raw_recipes_df['submitted'] = pd.to_datetime(raw_recipes_df['submitted'], format='%Y-%m-%d')\n",
    "\n",
    "# Print updated data types of the columns\n",
    "raw_recipes_df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "70751dc4",
   "metadata": {},
   "source": [
    "Let us now examine the data types within the user interactions dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "f7b58434",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 1132367 entries, 0 to 1132366\n",
      "Data columns (total 5 columns):\n",
      " #   Column     Non-Null Count    Dtype \n",
      "---  ------     --------------    ----- \n",
      " 0   user_id    1132367 non-null  int64 \n",
      " 1   recipe_id  1132367 non-null  int64 \n",
      " 2   date       1132367 non-null  object\n",
      " 3   rating     1132367 non-null  int64 \n",
      " 4   review     1132198 non-null  object\n",
      "dtypes: int64(3), object(2)\n",
      "memory usage: 43.2+ MB\n"
     ]
    }
   ],
   "source": [
    "# Info about the recipe dataset\n",
    "raw_users_df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5d40546a",
   "metadata": {},
   "source": [
    "Similar to the recipes dataset, the 'user_id' and 'recipe_id' columns are currently of `integer` data type, while the 'date' column is of `object` data type. To ensure consistency and facilitate data manipulation, we will convert the 'user_id' and 'recipe_id' columns to `object` data type. Additionally, we will convert the 'date' column to `datetime` data type for more convenient date-based operations."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "3e281103",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 1132367 entries, 0 to 1132366\n",
      "Data columns (total 5 columns):\n",
      " #   Column     Non-Null Count    Dtype         \n",
      "---  ------     --------------    -----         \n",
      " 0   user_id    1132367 non-null  object        \n",
      " 1   recipe_id  1132367 non-null  object        \n",
      " 2   date       1132367 non-null  datetime64[ns]\n",
      " 3   rating     1132367 non-null  int64         \n",
      " 4   review     1132198 non-null  object        \n",
      "dtypes: datetime64[ns](1), int64(1), object(3)\n",
      "memory usage: 43.2+ MB\n"
     ]
    }
   ],
   "source": [
    "# Convert 'user_id' and 'recipe_id' columns to object data type\n",
    "raw_users_df['user_id'] = raw_users_df['user_id'].astype(str)\n",
    "raw_users_df['recipe_id'] = raw_users_df['recipe_id'].astype(str)\n",
    "\n",
    "# Convert 'submitted' column to date data type\n",
    "raw_users_df['date'] = pd.to_datetime(raw_users_df['date'], format='%Y-%m-%d')\n",
    "\n",
    "# Print updated data types of the columns\n",
    "raw_users_df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bfc14724",
   "metadata": {},
   "source": [
    "### 3.2 Missing Data <a class=\"anchor\" id=\"threetwo\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fecf9fca",
   "metadata": {},
   "source": [
    "Now that we have gained a deeper understanding of the dataset, let us proceed to explore any potential missing values within it."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "5d421eba",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "name                 1\n",
       "id                   0\n",
       "minutes              0\n",
       "contributor_id       0\n",
       "submitted            0\n",
       "tags                 0\n",
       "nutrition            0\n",
       "n_steps              0\n",
       "steps                0\n",
       "description       4979\n",
       "ingredients          0\n",
       "n_ingredients        0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Checking for missing values\n",
    "raw_recipes_df.isna().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6e2c965e",
   "metadata": {},
   "source": [
    "There is 1 missing value in the 'name' column and 4,979 missing values in the 'description' column. values represent. Let us examine the specific rows that contain missing values to identify any discernible patterns and explore potential strategies for data imputation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "923005a3",
   "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>name</th>\n",
       "      <th>id</th>\n",
       "      <th>minutes</th>\n",
       "      <th>contributor_id</th>\n",
       "      <th>submitted</th>\n",
       "      <th>tags</th>\n",
       "      <th>nutrition</th>\n",
       "      <th>n_steps</th>\n",
       "      <th>steps</th>\n",
       "      <th>description</th>\n",
       "      <th>ingredients</th>\n",
       "      <th>n_ingredients</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>721</th>\n",
       "      <td>NaN</td>\n",
       "      <td>368257</td>\n",
       "      <td>10</td>\n",
       "      <td>779451</td>\n",
       "      <td>2009-04-27</td>\n",
       "      <td>['15-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>[1596.2, 249.0, 155.0, 0.0, 2.0, 112.0, 14.0]</td>\n",
       "      <td>6</td>\n",
       "      <td>['in a bowl , combine ingredients except for o...</td>\n",
       "      <td>-------------</td>\n",
       "      <td>['lemon', 'honey', 'horseradish mustard', 'gar...</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    name      id  minutes contributor_id  submitted  \\\n",
       "721  NaN  368257       10         779451 2009-04-27   \n",
       "\n",
       "                                                  tags  \\\n",
       "721  ['15-minutes-or-less', 'time-to-make', 'course...   \n",
       "\n",
       "                                         nutrition  n_steps  \\\n",
       "721  [1596.2, 249.0, 155.0, 0.0, 2.0, 112.0, 14.0]        6   \n",
       "\n",
       "                                                 steps    description  \\\n",
       "721  ['in a bowl , combine ingredients except for o...  -------------   \n",
       "\n",
       "                                           ingredients  n_ingredients  \n",
       "721  ['lemon', 'honey', 'horseradish mustard', 'gar...             10  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Row with missing name\n",
    "raw_recipes_df[raw_recipes_df['name'].isna()]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0e481bd1",
   "metadata": {},
   "source": [
    "It appears that the row with the missing 'name' value also lacks a corresponding description. Consquently, we are unable to infer a name from the description itself, as there is not textual information available to guide the imputation process. Considering that there is only one instance with a missing 'name' value, we can safely drop this row from the dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "e8539121",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Drop row with missing name\n",
    "raw_recipes_df.dropna(subset=['name'], inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "2b03f55f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "name                 0\n",
       "id                   0\n",
       "minutes              0\n",
       "contributor_id       0\n",
       "submitted            0\n",
       "tags                 0\n",
       "nutrition            0\n",
       "n_steps              0\n",
       "steps                0\n",
       "description       4979\n",
       "ingredients          0\n",
       "n_ingredients        0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Sanity check\n",
    "raw_recipes_df.isna().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "43343794",
   "metadata": {},
   "source": [
    "Let us take a look at the rows with a missing description."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "d4e5bef4",
   "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>name</th>\n",
       "      <th>id</th>\n",
       "      <th>minutes</th>\n",
       "      <th>contributor_id</th>\n",
       "      <th>submitted</th>\n",
       "      <th>tags</th>\n",
       "      <th>nutrition</th>\n",
       "      <th>n_steps</th>\n",
       "      <th>steps</th>\n",
       "      <th>description</th>\n",
       "      <th>ingredients</th>\n",
       "      <th>n_ingredients</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>212861</th>\n",
       "      <td>thermonuclear bbq shrimp</td>\n",
       "      <td>24214</td>\n",
       "      <td>72</td>\n",
       "      <td>23302</td>\n",
       "      <td>2002-04-02</td>\n",
       "      <td>['weeknight', 'time-to-make', 'course', 'main-...</td>\n",
       "      <td>[370.3, 24.0, 17.0, 14.0, 93.0, 40.0, 2.0]</td>\n",
       "      <td>7</td>\n",
       "      <td>['wash the shrimp , drain and place on skewers...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>['raw shrimp', 'habanero pepper', 'unsalted bu...</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3674</th>\n",
       "      <td>almond cookie bites</td>\n",
       "      <td>401761</td>\n",
       "      <td>16</td>\n",
       "      <td>1434673</td>\n",
       "      <td>2009-11-29</td>\n",
       "      <td>['30-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>[40.1, 3.0, 4.0, 1.0, 0.0, 8.0, 1.0]</td>\n",
       "      <td>16</td>\n",
       "      <td>['preheat oven to 350 degrees f', 'in medium b...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>['all-purpose flour', \"fisher chef's naturals ...</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>110086</th>\n",
       "      <td>hunter s lullaby</td>\n",
       "      <td>21351</td>\n",
       "      <td>70</td>\n",
       "      <td>32692</td>\n",
       "      <td>2002-03-03</td>\n",
       "      <td>['weeknight', 'time-to-make', 'course', 'main-...</td>\n",
       "      <td>[397.3, 42.0, 3.0, 4.0, 59.0, 36.0, 0.0]</td>\n",
       "      <td>10</td>\n",
       "      <td>['heat oil in deep frying pan', 'fry garlic un...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>['chicken', 'olive oil', 'garlic', 'fresh rose...</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35645</th>\n",
       "      <td>caramel pie</td>\n",
       "      <td>5057</td>\n",
       "      <td>0</td>\n",
       "      <td>1534</td>\n",
       "      <td>1999-11-29</td>\n",
       "      <td>['15-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>[518.6, 27.0, 264.0, 11.0, 14.0, 33.0, 27.0]</td>\n",
       "      <td>15</td>\n",
       "      <td>['sift together flour and 1 cup sugar', 'add e...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>['sugar', 'hot water', 'flour', 'egg yolks', '...</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51881</th>\n",
       "      <td>chocolate vinegar cake</td>\n",
       "      <td>7701</td>\n",
       "      <td>25</td>\n",
       "      <td>87236</td>\n",
       "      <td>2000-03-13</td>\n",
       "      <td>['30-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>[356.2, 16.0, 180.0, 15.0, 5.0, 33.0, 21.0]</td>\n",
       "      <td>16</td>\n",
       "      <td>['sift the flour , sugar , cocoa , baking soda...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>['flour', 'sugar', 'cocoa', 'baking soda', 'sa...</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>59992</th>\n",
       "      <td>crabby crock pot soup</td>\n",
       "      <td>13774</td>\n",
       "      <td>310</td>\n",
       "      <td>20754</td>\n",
       "      <td>2001-11-06</td>\n",
       "      <td>['weeknight', 'time-to-make', 'course', 'main-...</td>\n",
       "      <td>[381.3, 43.0, 2.0, 11.0, 17.0, 85.0, 6.0]</td>\n",
       "      <td>4</td>\n",
       "      <td>['combine all ingredients except sherry and cr...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>['crabmeat', 'milk', 'half-and-half', 'butter'...</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>219143</th>\n",
       "      <td>two mushroom quiche</td>\n",
       "      <td>53058</td>\n",
       "      <td>50</td>\n",
       "      <td>54716</td>\n",
       "      <td>2003-02-03</td>\n",
       "      <td>['60-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>[396.8, 47.0, 9.0, 26.0, 16.0, 74.0, 6.0]</td>\n",
       "      <td>10</td>\n",
       "      <td>['wash the mushrooms thoroughly', 'cut the mus...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>['pie shells', 'mushroom', 'wild mushroom', 's...</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77569</th>\n",
       "      <td>easy pecan pie</td>\n",
       "      <td>10487</td>\n",
       "      <td>65</td>\n",
       "      <td>12491</td>\n",
       "      <td>2001-07-29</td>\n",
       "      <td>['weeknight', 'time-to-make', 'course', 'main-...</td>\n",
       "      <td>[3909.2, 330.0, 923.0, 123.0, 81.0, 325.0, 164.0]</td>\n",
       "      <td>5</td>\n",
       "      <td>['preheat oven to 375', 'beat together all fil...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>['pie crusts', 'eggs', 'sugar', 'salt', 'butte...</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>153760</th>\n",
       "      <td>pasta primavera</td>\n",
       "      <td>7847</td>\n",
       "      <td>0</td>\n",
       "      <td>157425</td>\n",
       "      <td>2000-03-13</td>\n",
       "      <td>['15-minutes-or-less', 'time-to-make', 'main-i...</td>\n",
       "      <td>[786.9, 127.0, 7.0, 18.0, 12.0, 257.0, 2.0]</td>\n",
       "      <td>10</td>\n",
       "      <td>['melt the butter in a large sauce pan', 'add ...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>['salted butter', 'egg yolks', 'fresh black pe...</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68177</th>\n",
       "      <td>cumin dusted pork cutlets with citrus pan sauce</td>\n",
       "      <td>10399</td>\n",
       "      <td>20</td>\n",
       "      <td>8728</td>\n",
       "      <td>2001-07-24</td>\n",
       "      <td>['30-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>[236.5, 32.0, 11.0, 24.0, 3.0, 14.0, 4.0]</td>\n",
       "      <td>12</td>\n",
       "      <td>['mix first 4 ingredients on a plate', 'coat p...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>['all-purpose flour', 'ground cumin', 'salt', ...</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                   name      id  minutes  \\\n",
       "212861                         thermonuclear bbq shrimp   24214       72   \n",
       "3674                                almond cookie bites  401761       16   \n",
       "110086                                 hunter s lullaby   21351       70   \n",
       "35645                                       caramel pie    5057        0   \n",
       "51881                            chocolate vinegar cake    7701       25   \n",
       "59992                             crabby crock pot soup   13774      310   \n",
       "219143                              two mushroom quiche   53058       50   \n",
       "77569                                    easy pecan pie   10487       65   \n",
       "153760                                  pasta primavera    7847        0   \n",
       "68177   cumin dusted pork cutlets with citrus pan sauce   10399       20   \n",
       "\n",
       "       contributor_id  submitted  \\\n",
       "212861          23302 2002-04-02   \n",
       "3674          1434673 2009-11-29   \n",
       "110086          32692 2002-03-03   \n",
       "35645            1534 1999-11-29   \n",
       "51881           87236 2000-03-13   \n",
       "59992           20754 2001-11-06   \n",
       "219143          54716 2003-02-03   \n",
       "77569           12491 2001-07-29   \n",
       "153760         157425 2000-03-13   \n",
       "68177            8728 2001-07-24   \n",
       "\n",
       "                                                     tags  \\\n",
       "212861  ['weeknight', 'time-to-make', 'course', 'main-...   \n",
       "3674    ['30-minutes-or-less', 'time-to-make', 'course...   \n",
       "110086  ['weeknight', 'time-to-make', 'course', 'main-...   \n",
       "35645   ['15-minutes-or-less', 'time-to-make', 'course...   \n",
       "51881   ['30-minutes-or-less', 'time-to-make', 'course...   \n",
       "59992   ['weeknight', 'time-to-make', 'course', 'main-...   \n",
       "219143  ['60-minutes-or-less', 'time-to-make', 'course...   \n",
       "77569   ['weeknight', 'time-to-make', 'course', 'main-...   \n",
       "153760  ['15-minutes-or-less', 'time-to-make', 'main-i...   \n",
       "68177   ['30-minutes-or-less', 'time-to-make', 'course...   \n",
       "\n",
       "                                                nutrition  n_steps  \\\n",
       "212861         [370.3, 24.0, 17.0, 14.0, 93.0, 40.0, 2.0]        7   \n",
       "3674                 [40.1, 3.0, 4.0, 1.0, 0.0, 8.0, 1.0]       16   \n",
       "110086           [397.3, 42.0, 3.0, 4.0, 59.0, 36.0, 0.0]       10   \n",
       "35645        [518.6, 27.0, 264.0, 11.0, 14.0, 33.0, 27.0]       15   \n",
       "51881         [356.2, 16.0, 180.0, 15.0, 5.0, 33.0, 21.0]       16   \n",
       "59992           [381.3, 43.0, 2.0, 11.0, 17.0, 85.0, 6.0]        4   \n",
       "219143          [396.8, 47.0, 9.0, 26.0, 16.0, 74.0, 6.0]       10   \n",
       "77569   [3909.2, 330.0, 923.0, 123.0, 81.0, 325.0, 164.0]        5   \n",
       "153760        [786.9, 127.0, 7.0, 18.0, 12.0, 257.0, 2.0]       10   \n",
       "68177           [236.5, 32.0, 11.0, 24.0, 3.0, 14.0, 4.0]       12   \n",
       "\n",
       "                                                    steps description  \\\n",
       "212861  ['wash the shrimp , drain and place on skewers...         NaN   \n",
       "3674    ['preheat oven to 350 degrees f', 'in medium b...         NaN   \n",
       "110086  ['heat oil in deep frying pan', 'fry garlic un...         NaN   \n",
       "35645   ['sift together flour and 1 cup sugar', 'add e...         NaN   \n",
       "51881   ['sift the flour , sugar , cocoa , baking soda...         NaN   \n",
       "59992   ['combine all ingredients except sherry and cr...         NaN   \n",
       "219143  ['wash the mushrooms thoroughly', 'cut the mus...         NaN   \n",
       "77569   ['preheat oven to 375', 'beat together all fil...         NaN   \n",
       "153760  ['melt the butter in a large sauce pan', 'add ...         NaN   \n",
       "68177   ['mix first 4 ingredients on a plate', 'coat p...         NaN   \n",
       "\n",
       "                                              ingredients  n_ingredients  \n",
       "212861  ['raw shrimp', 'habanero pepper', 'unsalted bu...             13  \n",
       "3674    ['all-purpose flour', \"fisher chef's naturals ...              9  \n",
       "110086  ['chicken', 'olive oil', 'garlic', 'fresh rose...             10  \n",
       "35645   ['sugar', 'hot water', 'flour', 'egg yolks', '...             10  \n",
       "51881   ['flour', 'sugar', 'cocoa', 'baking soda', 'sa...             12  \n",
       "59992   ['crabmeat', 'milk', 'half-and-half', 'butter'...              9  \n",
       "219143  ['pie shells', 'mushroom', 'wild mushroom', 's...             12  \n",
       "77569   ['pie crusts', 'eggs', 'sugar', 'salt', 'butte...              7  \n",
       "153760  ['salted butter', 'egg yolks', 'fresh black pe...             10  \n",
       "68177   ['all-purpose flour', 'ground cumin', 'salt', ...             10  "
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Row with missing description\n",
    "raw_recipes_df[raw_recipes_df['description'].isna()].sample(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "360bd19e",
   "metadata": {},
   "source": [
    "Since the descriptions are written by the recipe contributors themselves and not available in the source, it is not feasible to impute missing values for the 'description' column. Let us look at the extent to which these missing values constitute the overall dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "a0fddc46",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Missing values make up 2.15 % of the overall data.\n"
     ]
    }
   ],
   "source": [
    "print('Missing values make up', round(raw_recipes_df.isna().sum().sum()/raw_recipes_df.shape[0]*100, 2), \"%\", \"of the overall data.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5bfe71dc",
   "metadata": {},
   "source": [
    "Given that the missing values in the 'description' column constitute a small subset of the overall dataset, we can safely drop these rows."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "b7ee1a59",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Drop rows with missing description\n",
    "raw_recipes_df.dropna(subset=['description'], inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "bf879405",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "name              0\n",
       "id                0\n",
       "minutes           0\n",
       "contributor_id    0\n",
       "submitted         0\n",
       "tags              0\n",
       "nutrition         0\n",
       "n_steps           0\n",
       "steps             0\n",
       "description       0\n",
       "ingredients       0\n",
       "n_ingredients     0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Sanity check\n",
    "raw_recipes_df.isna().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "29b09f5a",
   "metadata": {},
   "source": [
    "We will now take a look at the user interactions dataset to identify any missing values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "b6aeebe1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "user_id        0\n",
       "recipe_id      0\n",
       "date           0\n",
       "rating         0\n",
       "review       169\n",
       "dtype: int64"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "raw_users_df.isna().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1d16f62c",
   "metadata": {},
   "source": [
    "There are 169 missing values in the 'review' column. values represent. Similar to the 'description' column in the recipes dataset, as the reviews are written by users themselves and are not available in the source, it is not feasible to impute missing values for the 'review' column. Let us now assess the proportion of these missing values in relation to the overall dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "ad43a773",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Missing values make up 0.01 % of the overall data.\n"
     ]
    }
   ],
   "source": [
    "print('Missing values make up', round(raw_users_df.isna().sum().sum()/raw_users_df.shape[0]*100, 2), \"%\", \"of the overall data.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "656144a3",
   "metadata": {},
   "source": [
    "Given that the missing values in the 'review' column constitute a very small subset of the overall dataset, we can safely drop these rows."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "75adee50",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Drop rows with missing description\n",
    "raw_users_df.dropna(subset=['review'], inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "6d51797f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "user_id      0\n",
       "recipe_id    0\n",
       "date         0\n",
       "rating       0\n",
       "review       0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Sanity check\n",
    "raw_users_df.isna().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0c55be02",
   "metadata": {},
   "source": [
    "### 3.3 Duplicate Data <a class=\"anchor\" id=\"threethree\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a74366ad",
   "metadata": {},
   "source": [
    "Now that we have addressed the missing values, let us proceed to identify any potential duplicates within the datasets."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "d4f3d420",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "duplicated rows in recipes dataset: 0\n",
      "duplicated rows in user interactions dataset: 0\n"
     ]
    }
   ],
   "source": [
    "# Duplicated rows\n",
    "print(\"duplicated rows in recipes dataset:\", raw_recipes_df.duplicated().sum())\n",
    "print(\"duplicated rows in user interactions dataset:\", raw_users_df.duplicated().sum())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0c935593",
   "metadata": {},
   "source": [
    "Great! It appears that there are no duplicated rows within the datasets."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b77b95e4",
   "metadata": {},
   "source": [
    "### 3.4 Outliers <a class=\"anchor\" id=\"threefour\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "691beb57",
   "metadata": {},
   "source": [
    "In this section, we will examine the datasets to identify the presence of any outliers."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "915b4058",
   "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>minutes</th>\n",
       "      <th>n_steps</th>\n",
       "      <th>n_ingredients</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>2.266570e+05</td>\n",
       "      <td>226657.000000</td>\n",
       "      <td>226657.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>9.602472e+03</td>\n",
       "      <td>9.773927</td>\n",
       "      <td>9.054170</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>4.510715e+06</td>\n",
       "      <td>5.999824</td>\n",
       "      <td>3.740942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>2.000000e+01</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>6.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>4.000000e+01</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>9.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>6.500000e+01</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>11.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>2.147484e+09</td>\n",
       "      <td>145.000000</td>\n",
       "      <td>43.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            minutes        n_steps  n_ingredients\n",
       "count  2.266570e+05  226657.000000  226657.000000\n",
       "mean   9.602472e+03       9.773927       9.054170\n",
       "std    4.510715e+06       5.999824       3.740942\n",
       "min    0.000000e+00       0.000000       1.000000\n",
       "25%    2.000000e+01       6.000000       6.000000\n",
       "50%    4.000000e+01       9.000000       9.000000\n",
       "75%    6.500000e+01      12.000000      11.000000\n",
       "max    2.147484e+09     145.000000      43.000000"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Summary statistics of recipes dataset\n",
    "raw_recipes_df.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eb65f96b",
   "metadata": {},
   "source": [
    "The value \"2,147,484,000 minutes\" (max minutes) does indeed seem like an unusually long time to cook a recipe. The maximum value of \"2,147,484,000 minutes\" in the \"minutes\" column does indeed seem like an unusually long time to cook a recipe and certainly does appear to be an outlier. Similarly, the high number of steps (145) and ingredients (43) also seem to deviate from the average values. Let us examine these columns more closely to understand this further."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "6b5af7c5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x400 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "# Distribution\n",
    "# Create a figure with three subplots\n",
    "fig, ax = plt.subplots(1, 3, figsize=(15, 4))\n",
    "\n",
    "# Plot distribution of 'minutes' on the first subplot\n",
    "sns.distplot(raw_recipes_df[\"minutes\"], ax=ax[0])\n",
    "ax[0].set_xlabel('Minutes')\n",
    "ax[0].set_ylabel('Density')\n",
    "ax[0].set_title('Distribution of Cooking Time')\n",
    "\n",
    "# Plot distribution of 'n_steps' on the second subplot\n",
    "sns.distplot(raw_recipes_df[\"n_steps\"], ax=ax[1])\n",
    "ax[1].set_xlabel('Number of Steps')\n",
    "ax[1].set_ylabel('Density')\n",
    "ax[1].set_title('Distribution of Number of Steps')\n",
    "\n",
    "# Plot distribution of 'n_ingredients' on the third subplot\n",
    "sns.distplot(raw_recipes_df[\"n_ingredients\"], ax=ax[2])\n",
    "ax[2].set_xlabel('Number of Ingredients')\n",
    "ax[2].set_ylabel('Density')\n",
    "ax[2].set_title('Distribution of Number of Ingredients')\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "36e4967c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x400 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Boxplots\n",
    "# Create a figure with three subplots\n",
    "fig, ax = plt.subplots(1, 3, figsize=(15, 4))\n",
    "\n",
    "# Plot boxplot of 'minutes' on the first subplot\n",
    "sns.boxplot(data=raw_recipes_df[\"minutes\"], ax=ax[0])\n",
    "ax[0].set_ylabel('Minutes')\n",
    "ax[0].set_title('Distribution of Cooking Time')\n",
    "\n",
    "# Plot boxplot of 'n_steps' on the second subplot\n",
    "sns.boxplot(data=raw_recipes_df[\"n_steps\"], ax=ax[1])\n",
    "ax[1].set_ylabel('Number of Steps')\n",
    "ax[1].set_title('Distribution of Number of Steps')\n",
    "\n",
    "# Plot boxplot of 'n_ingredients' on the third subplot\n",
    "sns.boxplot(data=raw_recipes_df[\"n_ingredients\"], ax=ax[2])\n",
    "ax[2].set_ylabel('Number of Ingredients')\n",
    "ax[2].set_title('Distribution of Number of Ingredients')\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "084e7723",
   "metadata": {},
   "source": [
    "It appears that there may be outliers present in the dataset. To further investigate and identify these outliers, we will employ the Tukey method, a statistical technique commonly used to detect and analyze outliers."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "f1845c61",
   "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>name</th>\n",
       "      <th>id</th>\n",
       "      <th>minutes</th>\n",
       "      <th>contributor_id</th>\n",
       "      <th>submitted</th>\n",
       "      <th>tags</th>\n",
       "      <th>nutrition</th>\n",
       "      <th>n_steps</th>\n",
       "      <th>steps</th>\n",
       "      <th>description</th>\n",
       "      <th>ingredients</th>\n",
       "      <th>n_ingredients</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>206778</th>\n",
       "      <td>sweet rice pudding</td>\n",
       "      <td>454411</td>\n",
       "      <td>133</td>\n",
       "      <td>1889545</td>\n",
       "      <td>2011-04-25</td>\n",
       "      <td>['time-to-make', 'course', 'main-ingredient', ...</td>\n",
       "      <td>[338.6, 12.0, 185.0, 3.0, 16.0, 20.0, 19.0]</td>\n",
       "      <td>23</td>\n",
       "      <td>['place the rice in a medium-sized bowl', 'fil...</td>\n",
       "      <td>this is a traditional recipe from india and it...</td>\n",
       "      <td>['basmati rice', 'whole milk', 'granulated sug...</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>204465</th>\n",
       "      <td>super soft bread machine rolls with wheat germ</td>\n",
       "      <td>210241</td>\n",
       "      <td>133</td>\n",
       "      <td>260718</td>\n",
       "      <td>2007-02-09</td>\n",
       "      <td>['time-to-make', 'course', 'main-ingredient', ...</td>\n",
       "      <td>[1580.3, 93.0, 106.0, 92.0, 83.0, 171.0, 71.0]</td>\n",
       "      <td>7</td>\n",
       "      <td>['in bread machine pan , put in the following ...</td>\n",
       "      <td>this recipe is adapted from the buttery bread ...</td>\n",
       "      <td>['milk', 'vanilla essence', 'butter', 'sugar',...</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>94271</th>\n",
       "      <td>gooey butter cookies</td>\n",
       "      <td>34390</td>\n",
       "      <td>133</td>\n",
       "      <td>43121</td>\n",
       "      <td>2002-07-16</td>\n",
       "      <td>['weeknight', 'time-to-make', 'course', 'prepa...</td>\n",
       "      <td>[163.3, 14.0, 49.0, 8.0, 3.0, 23.0, 5.0]</td>\n",
       "      <td>8</td>\n",
       "      <td>['cream together cream cheese and butter', 'ad...</td>\n",
       "      <td>so easy to make and always a hit!</td>\n",
       "      <td>['cream cheese', 'butter', 'egg', 'vanilla', '...</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99018</th>\n",
       "      <td>grilled halibut with lemon tarragon mustard</td>\n",
       "      <td>116898</td>\n",
       "      <td>134</td>\n",
       "      <td>39547</td>\n",
       "      <td>2005-04-12</td>\n",
       "      <td>['time-to-make', 'course', 'main-ingredient', ...</td>\n",
       "      <td>[470.1, 23.0, 2.0, 16.0, 153.0, 12.0, 1.0]</td>\n",
       "      <td>8</td>\n",
       "      <td>['in a small bowl whisk together the lemon jui...</td>\n",
       "      <td>adapted from a recipe in \"the taste of summer\"...</td>\n",
       "      <td>['fresh lemon juice', 'lemon, zest of', 'dijon...</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>183119</th>\n",
       "      <td>sea bass with citrus and soy</td>\n",
       "      <td>23101</td>\n",
       "      <td>134</td>\n",
       "      <td>8688</td>\n",
       "      <td>2002-03-21</td>\n",
       "      <td>['weeknight', 'time-to-make', 'course', 'main-...</td>\n",
       "      <td>[275.1, 16.0, 24.0, 60.0, 69.0, 9.0, 3.0]</td>\n",
       "      <td>14</td>\n",
       "      <td>['mix first 6 ingredients in 8x8x2\" glass baki...</td>\n",
       "      <td>simple, elegant and healthy, too! what more ca...</td>\n",
       "      <td>['pineapple juice', 'orange juice', 'soy sauce...</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>106963</th>\n",
       "      <td>homemade vanilla</td>\n",
       "      <td>425681</td>\n",
       "      <td>259205</td>\n",
       "      <td>28177</td>\n",
       "      <td>2010-05-16</td>\n",
       "      <td>['time-to-make', 'preparation', '5-ingredients...</td>\n",
       "      <td>[69.4, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]</td>\n",
       "      <td>9</td>\n",
       "      <td>['slice the vanilla beans length-wise and scra...</td>\n",
       "      <td>found this recipe on tammy's blog (http://www....</td>\n",
       "      <td>['vanilla beans', 'vodka']</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38116</th>\n",
       "      <td>celtic druid s honey mead   meade   metheglin</td>\n",
       "      <td>216215</td>\n",
       "      <td>259260</td>\n",
       "      <td>242729</td>\n",
       "      <td>2007-03-11</td>\n",
       "      <td>['time-to-make', 'course', 'main-ingredient', ...</td>\n",
       "      <td>[531.0, 0.0, 561.0, 1.0, 1.0, 0.0, 47.0]</td>\n",
       "      <td>18</td>\n",
       "      <td>['equipement:', 'very large pan - to hold 12 p...</td>\n",
       "      <td>mead, made from fermented honey, was the earli...</td>\n",
       "      <td>['gingerroot', 'lemons, juice and rind of', 'w...</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>106563</th>\n",
       "      <td>homemade fruit liquers</td>\n",
       "      <td>291571</td>\n",
       "      <td>288000</td>\n",
       "      <td>553251</td>\n",
       "      <td>2008-03-12</td>\n",
       "      <td>['time-to-make', 'course', 'main-ingredient', ...</td>\n",
       "      <td>[836.2, 0.0, 333.0, 0.0, 0.0, 0.0, 27.0]</td>\n",
       "      <td>12</td>\n",
       "      <td>['rinse the fruit or berries , fruit must be c...</td>\n",
       "      <td>this should be a nice easy project for those w...</td>\n",
       "      <td>['berries', 'vodka', 'granulated sugar']</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>109624</th>\n",
       "      <td>how to preserve a husband</td>\n",
       "      <td>447963</td>\n",
       "      <td>1051200</td>\n",
       "      <td>576273</td>\n",
       "      <td>2011-02-01</td>\n",
       "      <td>['time-to-make', 'course', 'preparation', 'for...</td>\n",
       "      <td>[407.4, 57.0, 50.0, 1.0, 7.0, 115.0, 5.0]</td>\n",
       "      <td>9</td>\n",
       "      <td>['be careful in your selection', \"don't choose...</td>\n",
       "      <td>found this in a local wyoming cookbook \"a coll...</td>\n",
       "      <td>['cream', 'peach']</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>144074</th>\n",
       "      <td>no bake granola balls</td>\n",
       "      <td>261647</td>\n",
       "      <td>2147483647</td>\n",
       "      <td>464080</td>\n",
       "      <td>2007-10-26</td>\n",
       "      <td>['60-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>[330.3, 23.0, 110.0, 4.0, 15.0, 24.0, 15.0]</td>\n",
       "      <td>9</td>\n",
       "      <td>['preheat the oven to 350 degrees', 'spread oa...</td>\n",
       "      <td>healthy snacks that kids (and grown ups) will ...</td>\n",
       "      <td>['rolled oats', 'unsweetened dried shredded co...</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>23274 rows × 12 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                  name      id     minutes  \\\n",
       "206778                              sweet rice pudding  454411         133   \n",
       "204465  super soft bread machine rolls with wheat germ  210241         133   \n",
       "94271                             gooey butter cookies   34390         133   \n",
       "99018      grilled halibut with lemon tarragon mustard  116898         134   \n",
       "183119                    sea bass with citrus and soy   23101         134   \n",
       "...                                                ...     ...         ...   \n",
       "106963                                homemade vanilla  425681      259205   \n",
       "38116    celtic druid s honey mead   meade   metheglin  216215      259260   \n",
       "106563                          homemade fruit liquers  291571      288000   \n",
       "109624                       how to preserve a husband  447963     1051200   \n",
       "144074                           no bake granola balls  261647  2147483647   \n",
       "\n",
       "       contributor_id  submitted  \\\n",
       "206778        1889545 2011-04-25   \n",
       "204465         260718 2007-02-09   \n",
       "94271           43121 2002-07-16   \n",
       "99018           39547 2005-04-12   \n",
       "183119           8688 2002-03-21   \n",
       "...               ...        ...   \n",
       "106963          28177 2010-05-16   \n",
       "38116          242729 2007-03-11   \n",
       "106563         553251 2008-03-12   \n",
       "109624         576273 2011-02-01   \n",
       "144074         464080 2007-10-26   \n",
       "\n",
       "                                                     tags  \\\n",
       "206778  ['time-to-make', 'course', 'main-ingredient', ...   \n",
       "204465  ['time-to-make', 'course', 'main-ingredient', ...   \n",
       "94271   ['weeknight', 'time-to-make', 'course', 'prepa...   \n",
       "99018   ['time-to-make', 'course', 'main-ingredient', ...   \n",
       "183119  ['weeknight', 'time-to-make', 'course', 'main-...   \n",
       "...                                                   ...   \n",
       "106963  ['time-to-make', 'preparation', '5-ingredients...   \n",
       "38116   ['time-to-make', 'course', 'main-ingredient', ...   \n",
       "106563  ['time-to-make', 'course', 'main-ingredient', ...   \n",
       "109624  ['time-to-make', 'course', 'preparation', 'for...   \n",
       "144074  ['60-minutes-or-less', 'time-to-make', 'course...   \n",
       "\n",
       "                                             nutrition  n_steps  \\\n",
       "206778     [338.6, 12.0, 185.0, 3.0, 16.0, 20.0, 19.0]       23   \n",
       "204465  [1580.3, 93.0, 106.0, 92.0, 83.0, 171.0, 71.0]        7   \n",
       "94271         [163.3, 14.0, 49.0, 8.0, 3.0, 23.0, 5.0]        8   \n",
       "99018       [470.1, 23.0, 2.0, 16.0, 153.0, 12.0, 1.0]        8   \n",
       "183119       [275.1, 16.0, 24.0, 60.0, 69.0, 9.0, 3.0]       14   \n",
       "...                                                ...      ...   \n",
       "106963            [69.4, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]        9   \n",
       "38116         [531.0, 0.0, 561.0, 1.0, 1.0, 0.0, 47.0]       18   \n",
       "106563        [836.2, 0.0, 333.0, 0.0, 0.0, 0.0, 27.0]       12   \n",
       "109624       [407.4, 57.0, 50.0, 1.0, 7.0, 115.0, 5.0]        9   \n",
       "144074     [330.3, 23.0, 110.0, 4.0, 15.0, 24.0, 15.0]        9   \n",
       "\n",
       "                                                    steps  \\\n",
       "206778  ['place the rice in a medium-sized bowl', 'fil...   \n",
       "204465  ['in bread machine pan , put in the following ...   \n",
       "94271   ['cream together cream cheese and butter', 'ad...   \n",
       "99018   ['in a small bowl whisk together the lemon jui...   \n",
       "183119  ['mix first 6 ingredients in 8x8x2\" glass baki...   \n",
       "...                                                   ...   \n",
       "106963  ['slice the vanilla beans length-wise and scra...   \n",
       "38116   ['equipement:', 'very large pan - to hold 12 p...   \n",
       "106563  ['rinse the fruit or berries , fruit must be c...   \n",
       "109624  ['be careful in your selection', \"don't choose...   \n",
       "144074  ['preheat the oven to 350 degrees', 'spread oa...   \n",
       "\n",
       "                                              description  \\\n",
       "206778  this is a traditional recipe from india and it...   \n",
       "204465  this recipe is adapted from the buttery bread ...   \n",
       "94271                   so easy to make and always a hit!   \n",
       "99018   adapted from a recipe in \"the taste of summer\"...   \n",
       "183119  simple, elegant and healthy, too! what more ca...   \n",
       "...                                                   ...   \n",
       "106963  found this recipe on tammy's blog (http://www....   \n",
       "38116   mead, made from fermented honey, was the earli...   \n",
       "106563  this should be a nice easy project for those w...   \n",
       "109624  found this in a local wyoming cookbook \"a coll...   \n",
       "144074  healthy snacks that kids (and grown ups) will ...   \n",
       "\n",
       "                                              ingredients  n_ingredients  \n",
       "206778  ['basmati rice', 'whole milk', 'granulated sug...              7  \n",
       "204465  ['milk', 'vanilla essence', 'butter', 'sugar',...              9  \n",
       "94271   ['cream cheese', 'butter', 'egg', 'vanilla', '...              6  \n",
       "99018   ['fresh lemon juice', 'lemon, zest of', 'dijon...              9  \n",
       "183119  ['pineapple juice', 'orange juice', 'soy sauce...              8  \n",
       "...                                                   ...            ...  \n",
       "106963                         ['vanilla beans', 'vodka']              2  \n",
       "38116   ['gingerroot', 'lemons, juice and rind of', 'w...              7  \n",
       "106563           ['berries', 'vodka', 'granulated sugar']              3  \n",
       "109624                                 ['cream', 'peach']              2  \n",
       "144074  ['rolled oats', 'unsweetened dried shredded co...              8  \n",
       "\n",
       "[23274 rows x 12 columns]"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Cooking time\n",
    "# Calculate the first quartile, third quartile and the interquartile range\n",
    "Q1 = raw_recipes_df['minutes'].quantile(0.25)\n",
    "Q3 = raw_recipes_df['minutes'].quantile(0.75)\n",
    "IQR = Q3 - Q1\n",
    "\n",
    "# Calculate the maximum value and minimum values according to the Tukey rule\n",
    "max_value = Q3 + 1.5 * IQR\n",
    "min_value = Q1 - 1.5 * IQR\n",
    "\n",
    "# Filter the data for values that are greater than max_value or less than min_value\n",
    "minutes_outliers = raw_recipes_df[(raw_recipes_df['minutes'] > max_value) | (raw_recipes_df['minutes'] < min_value)]\n",
    "minutes_outliers.sort_values('minutes')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f20411cd",
   "metadata": {},
   "source": [
    "As observed from the results, the application of the Tukey method for outlier detection tends to filter out a considerable number of recipes that may have longer preparation times but are still reasonable. For instance, recipes like pickles, extracts, and liqueurs often require several days or even weeks to prepare, and these longer durations should not be considered as outliers in the context of their specific culinary processes. "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "32a1a19c",
   "metadata": {},
   "source": [
    "The dataset includes an extreme outlier with a preparation time of 1,051,200 minutes, which corresponds to the recipe '[No Bake Granola Balls](https://www.food.com/recipe/no-bake-granola-balls-261647)'. Upon further investigation, it appears that this value may be the result of a data collection error or scraping issue. To address this, we will manually input the correct preparation time for the 'No Bake Granola Balls' recipe based on reliable sources. While this recipe might possess unique value or significance, for the purpose of this exploration, we will exclude it from further analysis. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "6a3544f5",
   "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>name</th>\n",
       "      <th>id</th>\n",
       "      <th>minutes</th>\n",
       "      <th>contributor_id</th>\n",
       "      <th>submitted</th>\n",
       "      <th>tags</th>\n",
       "      <th>nutrition</th>\n",
       "      <th>n_steps</th>\n",
       "      <th>steps</th>\n",
       "      <th>description</th>\n",
       "      <th>ingredients</th>\n",
       "      <th>n_ingredients</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>144074</th>\n",
       "      <td>no bake granola balls</td>\n",
       "      <td>261647</td>\n",
       "      <td>25</td>\n",
       "      <td>464080</td>\n",
       "      <td>2007-10-26</td>\n",
       "      <td>['60-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>[330.3, 23.0, 110.0, 4.0, 15.0, 24.0, 15.0]</td>\n",
       "      <td>9</td>\n",
       "      <td>['preheat the oven to 350 degrees', 'spread oa...</td>\n",
       "      <td>healthy snacks that kids (and grown ups) will ...</td>\n",
       "      <td>['rolled oats', 'unsweetened dried shredded co...</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                         name      id  minutes contributor_id  submitted  \\\n",
       "144074  no bake granola balls  261647       25         464080 2007-10-26   \n",
       "\n",
       "                                                     tags  \\\n",
       "144074  ['60-minutes-or-less', 'time-to-make', 'course...   \n",
       "\n",
       "                                          nutrition  n_steps  \\\n",
       "144074  [330.3, 23.0, 110.0, 4.0, 15.0, 24.0, 15.0]        9   \n",
       "\n",
       "                                                    steps  \\\n",
       "144074  ['preheat the oven to 350 degrees', 'spread oa...   \n",
       "\n",
       "                                              description  \\\n",
       "144074  healthy snacks that kids (and grown ups) will ...   \n",
       "\n",
       "                                              ingredients  n_ingredients  \n",
       "144074  ['rolled oats', 'unsweetened dried shredded co...              8  "
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Replace thie cooking time with 25 minutes\n",
    "raw_recipes_df.loc[raw_recipes_df['id'] == '261647', 'minutes'] = 25\n",
    "\n",
    "# Check value\n",
    "raw_recipes_df[raw_recipes_df['id'] == '261647']"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "41d0c6c5",
   "metadata": {},
   "source": [
    "As for the second extreme outlier with a preparation time of 1,051,200 minutes, attributed to the '[How to Preserve a Husband](https://www.food.com/recipe/how-to-preserve-a-husband-447963)'recipe, while it may have its unique significance, we will exclude it from further exploration and analysis to focus on the majority of recipes in the dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "08268cc9",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Remove row\n",
    "raw_recipes_df = raw_recipes_df[raw_recipes_df['id'] != '447963']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "d79f651c",
   "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>name</th>\n",
       "      <th>id</th>\n",
       "      <th>minutes</th>\n",
       "      <th>contributor_id</th>\n",
       "      <th>submitted</th>\n",
       "      <th>tags</th>\n",
       "      <th>nutrition</th>\n",
       "      <th>n_steps</th>\n",
       "      <th>steps</th>\n",
       "      <th>description</th>\n",
       "      <th>ingredients</th>\n",
       "      <th>n_ingredients</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5497</th>\n",
       "      <td>angelica liqueur</td>\n",
       "      <td>70551</td>\n",
       "      <td>201610</td>\n",
       "      <td>59064</td>\n",
       "      <td>2003-09-04</td>\n",
       "      <td>['weeknight', 'time-to-make', 'course', 'cuisi...</td>\n",
       "      <td>[891.8, 0.0, 906.0, 0.0, 1.0, 0.0, 76.0]</td>\n",
       "      <td>7</td>\n",
       "      <td>['crush the fennel , aniseed , coriander seeds...</td>\n",
       "      <td>finish off a large meal with a tiny glass of t...</td>\n",
       "      <td>['fennel seed', 'anise seed', 'coriander seeds...</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>155044</th>\n",
       "      <td>peach brandy</td>\n",
       "      <td>98912</td>\n",
       "      <td>216015</td>\n",
       "      <td>34146</td>\n",
       "      <td>2004-09-01</td>\n",
       "      <td>['time-to-make', 'course', 'main-ingredient', ...</td>\n",
       "      <td>[3645.3, 3.0, 2040.0, 0.0, 16.0, 0.0, 173.0]</td>\n",
       "      <td>9</td>\n",
       "      <td>['use only fully ripened , unblemished fruit',...</td>\n",
       "      <td>posted in reply to a message board request for</td>\n",
       "      <td>['fresh peaches', 'granulated sugar', 'honey',...</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>106963</th>\n",
       "      <td>homemade vanilla</td>\n",
       "      <td>425681</td>\n",
       "      <td>259205</td>\n",
       "      <td>28177</td>\n",
       "      <td>2010-05-16</td>\n",
       "      <td>['time-to-make', 'preparation', '5-ingredients...</td>\n",
       "      <td>[69.4, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]</td>\n",
       "      <td>9</td>\n",
       "      <td>['slice the vanilla beans length-wise and scra...</td>\n",
       "      <td>found this recipe on tammy's blog (http://www....</td>\n",
       "      <td>['vanilla beans', 'vodka']</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38116</th>\n",
       "      <td>celtic druid s honey mead   meade   metheglin</td>\n",
       "      <td>216215</td>\n",
       "      <td>259260</td>\n",
       "      <td>242729</td>\n",
       "      <td>2007-03-11</td>\n",
       "      <td>['time-to-make', 'course', 'main-ingredient', ...</td>\n",
       "      <td>[531.0, 0.0, 561.0, 1.0, 1.0, 0.0, 47.0]</td>\n",
       "      <td>18</td>\n",
       "      <td>['equipement:', 'very large pan - to hold 12 p...</td>\n",
       "      <td>mead, made from fermented honey, was the earli...</td>\n",
       "      <td>['gingerroot', 'lemons, juice and rind of', 'w...</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>106563</th>\n",
       "      <td>homemade fruit liquers</td>\n",
       "      <td>291571</td>\n",
       "      <td>288000</td>\n",
       "      <td>553251</td>\n",
       "      <td>2008-03-12</td>\n",
       "      <td>['time-to-make', 'course', 'main-ingredient', ...</td>\n",
       "      <td>[836.2, 0.0, 333.0, 0.0, 0.0, 0.0, 27.0]</td>\n",
       "      <td>12</td>\n",
       "      <td>['rinse the fruit or berries , fruit must be c...</td>\n",
       "      <td>this should be a nice easy project for those w...</td>\n",
       "      <td>['berries', 'vodka', 'granulated sugar']</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                 name      id  minutes  \\\n",
       "5497                                 angelica liqueur   70551   201610   \n",
       "155044                                   peach brandy   98912   216015   \n",
       "106963                               homemade vanilla  425681   259205   \n",
       "38116   celtic druid s honey mead   meade   metheglin  216215   259260   \n",
       "106563                         homemade fruit liquers  291571   288000   \n",
       "\n",
       "       contributor_id  submitted  \\\n",
       "5497            59064 2003-09-04   \n",
       "155044          34146 2004-09-01   \n",
       "106963          28177 2010-05-16   \n",
       "38116          242729 2007-03-11   \n",
       "106563         553251 2008-03-12   \n",
       "\n",
       "                                                     tags  \\\n",
       "5497    ['weeknight', 'time-to-make', 'course', 'cuisi...   \n",
       "155044  ['time-to-make', 'course', 'main-ingredient', ...   \n",
       "106963  ['time-to-make', 'preparation', '5-ingredients...   \n",
       "38116   ['time-to-make', 'course', 'main-ingredient', ...   \n",
       "106563  ['time-to-make', 'course', 'main-ingredient', ...   \n",
       "\n",
       "                                           nutrition  n_steps  \\\n",
       "5497        [891.8, 0.0, 906.0, 0.0, 1.0, 0.0, 76.0]        7   \n",
       "155044  [3645.3, 3.0, 2040.0, 0.0, 16.0, 0.0, 173.0]        9   \n",
       "106963          [69.4, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]        9   \n",
       "38116       [531.0, 0.0, 561.0, 1.0, 1.0, 0.0, 47.0]       18   \n",
       "106563      [836.2, 0.0, 333.0, 0.0, 0.0, 0.0, 27.0]       12   \n",
       "\n",
       "                                                    steps  \\\n",
       "5497    ['crush the fennel , aniseed , coriander seeds...   \n",
       "155044  ['use only fully ripened , unblemished fruit',...   \n",
       "106963  ['slice the vanilla beans length-wise and scra...   \n",
       "38116   ['equipement:', 'very large pan - to hold 12 p...   \n",
       "106563  ['rinse the fruit or berries , fruit must be c...   \n",
       "\n",
       "                                              description  \\\n",
       "5497    finish off a large meal with a tiny glass of t...   \n",
       "155044    posted in reply to a message board request for    \n",
       "106963  found this recipe on tammy's blog (http://www....   \n",
       "38116   mead, made from fermented honey, was the earli...   \n",
       "106563  this should be a nice easy project for those w...   \n",
       "\n",
       "                                              ingredients  n_ingredients  \n",
       "5497    ['fennel seed', 'anise seed', 'coriander seeds...              7  \n",
       "155044  ['fresh peaches', 'granulated sugar', 'honey',...              4  \n",
       "106963                         ['vanilla beans', 'vodka']              2  \n",
       "38116   ['gingerroot', 'lemons, juice and rind of', 'w...              7  \n",
       "106563           ['berries', 'vodka', 'granulated sugar']              3  "
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Sanity check\n",
    "raw_recipes_df.sort_values('minutes', ascending=True).tail()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c76ffd09",
   "metadata": {},
   "source": [
    "Having addressed the outliers in cooking time, let us now shift our focus to the number of steps involved in each recipe."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "ef301f33",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize boxen plot\n",
    "sns.set(style = \"whitegrid\")\n",
    "ax = sns.boxenplot(x = raw_recipes_df['n_steps'])\n",
    "ax.set_xlabel('Number of steps per recipe')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "3dc6468f",
   "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>name</th>\n",
       "      <th>id</th>\n",
       "      <th>minutes</th>\n",
       "      <th>contributor_id</th>\n",
       "      <th>submitted</th>\n",
       "      <th>tags</th>\n",
       "      <th>nutrition</th>\n",
       "      <th>n_steps</th>\n",
       "      <th>steps</th>\n",
       "      <th>description</th>\n",
       "      <th>ingredients</th>\n",
       "      <th>n_ingredients</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>216131</th>\n",
       "      <td>traditional iraqi  casserole   tepsi baytinijan</td>\n",
       "      <td>288850</td>\n",
       "      <td>90</td>\n",
       "      <td>510313</td>\n",
       "      <td>2008-02-27</td>\n",
       "      <td>['time-to-make', 'course', 'main-ingredient', ...</td>\n",
       "      <td>[375.5, 20.0, 51.0, 7.0, 44.0, 25.0, 15.0]</td>\n",
       "      <td>22</td>\n",
       "      <td>['peel the eggplant in wide stripes and remove...</td>\n",
       "      <td>in iraq tepsi baytinijan is a very famous dish...</td>\n",
       "      <td>['eggplants', 'tomatoes', 'onion', 'garlic clo...</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70506</th>\n",
       "      <td>deep dish pizza with sausage  garlic  and mozz...</td>\n",
       "      <td>224394</td>\n",
       "      <td>75</td>\n",
       "      <td>37779</td>\n",
       "      <td>2007-04-24</td>\n",
       "      <td>['time-to-make', 'course', 'main-ingredient', ...</td>\n",
       "      <td>[929.5, 104.0, 48.0, 113.0, 107.0, 147.0, 7.0]</td>\n",
       "      <td>22</td>\n",
       "      <td>['position oven rack on the 2nd-lowest level i...</td>\n",
       "      <td>a husband favorite</td>\n",
       "      <td>['olive oil', 'pizza dough', 'mild italian sau...</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>169411</th>\n",
       "      <td>quick tikka masala</td>\n",
       "      <td>369865</td>\n",
       "      <td>45</td>\n",
       "      <td>31807</td>\n",
       "      <td>2009-05-05</td>\n",
       "      <td>['60-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>[514.5, 41.0, 8.0, 8.0, 114.0, 54.0, 2.0]</td>\n",
       "      <td>22</td>\n",
       "      <td>['grate 1 piece of ginger into a large zip top...</td>\n",
       "      <td>adapted from blake royer's \"dinner tonight\" co...</td>\n",
       "      <td>['fresh ginger', 'garlic clove', 'salt', 'grou...</td>\n",
       "      <td>17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33105</th>\n",
       "      <td>cabbage tart</td>\n",
       "      <td>62295</td>\n",
       "      <td>85</td>\n",
       "      <td>80353</td>\n",
       "      <td>2003-05-15</td>\n",
       "      <td>['weeknight', 'time-to-make', 'course', 'main-...</td>\n",
       "      <td>[614.3, 65.0, 18.0, 41.0, 29.0, 124.0, 14.0]</td>\n",
       "      <td>22</td>\n",
       "      <td>['for pastry: in a large bowl , combine flour ...</td>\n",
       "      <td>kind of like a cabbage quiche. very tasty - do...</td>\n",
       "      <td>['all-purpose flour', 'salt', 'butter', 'egg y...</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>132832</th>\n",
       "      <td>meatloaf with tomato gravy</td>\n",
       "      <td>112892</td>\n",
       "      <td>90</td>\n",
       "      <td>126440</td>\n",
       "      <td>2005-03-07</td>\n",
       "      <td>['time-to-make', 'course', 'main-ingredient', ...</td>\n",
       "      <td>[771.1, 84.0, 26.0, 89.0, 93.0, 97.0, 7.0]</td>\n",
       "      <td>22</td>\n",
       "      <td>['preheat the oven to 350f', 'in a large skill...</td>\n",
       "      <td>this recipe came from food network. i have mad...</td>\n",
       "      <td>['dry breadcrumbs', 'olive oil', 'onion', 'gar...</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70088</th>\n",
       "      <td>death by chocolate a la trellis</td>\n",
       "      <td>356724</td>\n",
       "      <td>1680</td>\n",
       "      <td>467583</td>\n",
       "      <td>2009-02-19</td>\n",
       "      <td>['time-to-make', 'course', 'main-ingredient', ...</td>\n",
       "      <td>[1460.8, 189.0, 262.0, 14.0, 50.0, 380.0, 36.0]</td>\n",
       "      <td>100</td>\n",
       "      <td>['take a deep breath', \"make sure that you wil...</td>\n",
       "      <td>warning: this dessert is not meant to be eaten...</td>\n",
       "      <td>['egg whites', 'cream of tartar', 'salt', 'gra...</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1627</th>\n",
       "      <td>5 cheese crab lasagna with roasted garlic and ...</td>\n",
       "      <td>65007</td>\n",
       "      <td>245</td>\n",
       "      <td>87300</td>\n",
       "      <td>2003-06-18</td>\n",
       "      <td>['weeknight', 'time-to-make', 'course', 'main-...</td>\n",
       "      <td>[442.8, 35.0, 27.0, 33.0, 50.0, 60.0, 10.0]</td>\n",
       "      <td>108</td>\n",
       "      <td>['roast garlic: place oven rack on second notc...</td>\n",
       "      <td>this creamy, gooey, gorgeous white lasagna wil...</td>\n",
       "      <td>['garlic', 'extra virgin olive oil', 'dry whit...</td>\n",
       "      <td>24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51532</th>\n",
       "      <td>chocolate raspberry ruffle cake</td>\n",
       "      <td>156874</td>\n",
       "      <td>1080</td>\n",
       "      <td>10033</td>\n",
       "      <td>2006-02-21</td>\n",
       "      <td>['time-to-make', 'course', 'main-ingredient', ...</td>\n",
       "      <td>[5080.8, 616.0, 967.0, 25.0, 147.0, 1211.0, 12...</td>\n",
       "      <td>110</td>\n",
       "      <td>['first of all: these are not typical directio...</td>\n",
       "      <td>alice medrich recipe, found in baking with jul...</td>\n",
       "      <td>['unsalted butter', 'pure vanilla extract', 'a...</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51528</th>\n",
       "      <td>chocolate raspberry mousse truffle cake   rasp...</td>\n",
       "      <td>192468</td>\n",
       "      <td>240</td>\n",
       "      <td>232669</td>\n",
       "      <td>2006-10-27</td>\n",
       "      <td>['time-to-make', 'course', 'main-ingredient', ...</td>\n",
       "      <td>[697.4, 57.0, 249.0, 9.0, 13.0, 105.0, 29.0]</td>\n",
       "      <td>113</td>\n",
       "      <td>['make the chocolate sour cream cake:', 'prehe...</td>\n",
       "      <td>this recipe is the winner, hands down! just in...</td>\n",
       "      <td>['cake flour', 'baking soda', 'salt', 'godiva ...</td>\n",
       "      <td>24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78633</th>\n",
       "      <td>easy yeast bread with variations</td>\n",
       "      <td>158120</td>\n",
       "      <td>210</td>\n",
       "      <td>182358</td>\n",
       "      <td>2006-02-28</td>\n",
       "      <td>['time-to-make', 'course', 'preparation', 'hea...</td>\n",
       "      <td>[1784.5, 62.0, 104.0, 151.0, 103.0, 42.0, 99.0]</td>\n",
       "      <td>145</td>\n",
       "      <td>['white bread:', 'lightly spooning flour into ...</td>\n",
       "      <td>found online in response to a request and now ...</td>\n",
       "      <td>['all-purpose flour', 'dry yeast', 'sugar', 's...</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>9837 rows × 12 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                     name      id  minutes  \\\n",
       "216131    traditional iraqi  casserole   tepsi baytinijan  288850       90   \n",
       "70506   deep dish pizza with sausage  garlic  and mozz...  224394       75   \n",
       "169411                                 quick tikka masala  369865       45   \n",
       "33105                                        cabbage tart   62295       85   \n",
       "132832                         meatloaf with tomato gravy  112892       90   \n",
       "...                                                   ...     ...      ...   \n",
       "70088                     death by chocolate a la trellis  356724     1680   \n",
       "1627    5 cheese crab lasagna with roasted garlic and ...   65007      245   \n",
       "51532                     chocolate raspberry ruffle cake  156874     1080   \n",
       "51528   chocolate raspberry mousse truffle cake   rasp...  192468      240   \n",
       "78633                    easy yeast bread with variations  158120      210   \n",
       "\n",
       "       contributor_id  submitted  \\\n",
       "216131         510313 2008-02-27   \n",
       "70506           37779 2007-04-24   \n",
       "169411          31807 2009-05-05   \n",
       "33105           80353 2003-05-15   \n",
       "132832         126440 2005-03-07   \n",
       "...               ...        ...   \n",
       "70088          467583 2009-02-19   \n",
       "1627            87300 2003-06-18   \n",
       "51532           10033 2006-02-21   \n",
       "51528          232669 2006-10-27   \n",
       "78633          182358 2006-02-28   \n",
       "\n",
       "                                                     tags  \\\n",
       "216131  ['time-to-make', 'course', 'main-ingredient', ...   \n",
       "70506   ['time-to-make', 'course', 'main-ingredient', ...   \n",
       "169411  ['60-minutes-or-less', 'time-to-make', 'course...   \n",
       "33105   ['weeknight', 'time-to-make', 'course', 'main-...   \n",
       "132832  ['time-to-make', 'course', 'main-ingredient', ...   \n",
       "...                                                   ...   \n",
       "70088   ['time-to-make', 'course', 'main-ingredient', ...   \n",
       "1627    ['weeknight', 'time-to-make', 'course', 'main-...   \n",
       "51532   ['time-to-make', 'course', 'main-ingredient', ...   \n",
       "51528   ['time-to-make', 'course', 'main-ingredient', ...   \n",
       "78633   ['time-to-make', 'course', 'preparation', 'hea...   \n",
       "\n",
       "                                                nutrition  n_steps  \\\n",
       "216131         [375.5, 20.0, 51.0, 7.0, 44.0, 25.0, 15.0]       22   \n",
       "70506      [929.5, 104.0, 48.0, 113.0, 107.0, 147.0, 7.0]       22   \n",
       "169411          [514.5, 41.0, 8.0, 8.0, 114.0, 54.0, 2.0]       22   \n",
       "33105        [614.3, 65.0, 18.0, 41.0, 29.0, 124.0, 14.0]       22   \n",
       "132832         [771.1, 84.0, 26.0, 89.0, 93.0, 97.0, 7.0]       22   \n",
       "...                                                   ...      ...   \n",
       "70088     [1460.8, 189.0, 262.0, 14.0, 50.0, 380.0, 36.0]      100   \n",
       "1627          [442.8, 35.0, 27.0, 33.0, 50.0, 60.0, 10.0]      108   \n",
       "51532   [5080.8, 616.0, 967.0, 25.0, 147.0, 1211.0, 12...      110   \n",
       "51528        [697.4, 57.0, 249.0, 9.0, 13.0, 105.0, 29.0]      113   \n",
       "78633     [1784.5, 62.0, 104.0, 151.0, 103.0, 42.0, 99.0]      145   \n",
       "\n",
       "                                                    steps  \\\n",
       "216131  ['peel the eggplant in wide stripes and remove...   \n",
       "70506   ['position oven rack on the 2nd-lowest level i...   \n",
       "169411  ['grate 1 piece of ginger into a large zip top...   \n",
       "33105   ['for pastry: in a large bowl , combine flour ...   \n",
       "132832  ['preheat the oven to 350f', 'in a large skill...   \n",
       "...                                                   ...   \n",
       "70088   ['take a deep breath', \"make sure that you wil...   \n",
       "1627    ['roast garlic: place oven rack on second notc...   \n",
       "51532   ['first of all: these are not typical directio...   \n",
       "51528   ['make the chocolate sour cream cake:', 'prehe...   \n",
       "78633   ['white bread:', 'lightly spooning flour into ...   \n",
       "\n",
       "                                              description  \\\n",
       "216131  in iraq tepsi baytinijan is a very famous dish...   \n",
       "70506                                  a husband favorite   \n",
       "169411  adapted from blake royer's \"dinner tonight\" co...   \n",
       "33105   kind of like a cabbage quiche. very tasty - do...   \n",
       "132832  this recipe came from food network. i have mad...   \n",
       "...                                                   ...   \n",
       "70088   warning: this dessert is not meant to be eaten...   \n",
       "1627    this creamy, gooey, gorgeous white lasagna wil...   \n",
       "51532   alice medrich recipe, found in baking with jul...   \n",
       "51528   this recipe is the winner, hands down! just in...   \n",
       "78633   found online in response to a request and now ...   \n",
       "\n",
       "                                              ingredients  n_ingredients  \n",
       "216131  ['eggplants', 'tomatoes', 'onion', 'garlic clo...             10  \n",
       "70506   ['olive oil', 'pizza dough', 'mild italian sau...              7  \n",
       "169411  ['fresh ginger', 'garlic clove', 'salt', 'grou...             17  \n",
       "33105   ['all-purpose flour', 'salt', 'butter', 'egg y...             12  \n",
       "132832  ['dry breadcrumbs', 'olive oil', 'onion', 'gar...             19  \n",
       "...                                                   ...            ...  \n",
       "70088   ['egg whites', 'cream of tartar', 'salt', 'gra...             19  \n",
       "1627    ['garlic', 'extra virgin olive oil', 'dry whit...             24  \n",
       "51532   ['unsalted butter', 'pure vanilla extract', 'a...             13  \n",
       "51528   ['cake flour', 'baking soda', 'salt', 'godiva ...             24  \n",
       "78633   ['all-purpose flour', 'dry yeast', 'sugar', 's...              8  \n",
       "\n",
       "[9837 rows x 12 columns]"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Number of steps\n",
    "# Calculate the first quartile, third quartile and the interquartile range\n",
    "Q1 = raw_recipes_df['n_steps'].quantile(0.25)\n",
    "Q3 = raw_recipes_df['n_steps'].quantile(0.75)\n",
    "IQR = Q3 - Q1\n",
    "\n",
    "# Calculate the maximum value and minimum values according to the Tukey rule\n",
    "max_value = Q3 + 1.5 * IQR\n",
    "min_value = Q1 - 1.5 * IQR\n",
    "\n",
    "# Filter the data for values that are greater than max_value or less than min_value\n",
    "nsteps_outliers = raw_recipes_df[(raw_recipes_df['n_steps'] > max_value) | (raw_recipes_df['n_steps'] < min_value)]\n",
    "nsteps_outliers.sort_values('n_steps')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "5f0b5873",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Based on the Tukey test, 9837 rows are outliers.\n"
     ]
    }
   ],
   "source": [
    "# Print the number of outliers\n",
    "print(f\"Based on the Tukey test, {len(nsteps_outliers)} rows are outliers.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ba9e76ab",
   "metadata": {},
   "source": [
    "To ensure data integrity and maintain the reliability of our analysis, we will proceed to remove these outlier rows from the dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "7f5b3447",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Drop outliers\n",
    "raw_recipes_df = raw_recipes_df.drop(nsteps_outliers.index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "14528baa",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize boxen plot\n",
    "sns.set(style = \"whitegrid\")\n",
    "ax = sns.boxenplot(x = raw_recipes_df['n_steps'])\n",
    "ax.set_xlabel('Number of steps per recipe')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "efd9343e",
   "metadata": {},
   "source": [
    "Looking at the boxen plot, there seems to be an outlier on the lower end of the spectrum. To ensure the validity of our analysis, we will retain only the recipes that have at least one step. Since each recipe should have a minimum of one step, it is important to exclude any recipes that do not meet this requirement. By keeping only the recipes with more than zero steps, we ensure that our dataset consists of complete and meaningful recipe entries."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "3641da71",
   "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>name</th>\n",
       "      <th>id</th>\n",
       "      <th>minutes</th>\n",
       "      <th>contributor_id</th>\n",
       "      <th>submitted</th>\n",
       "      <th>tags</th>\n",
       "      <th>nutrition</th>\n",
       "      <th>n_steps</th>\n",
       "      <th>steps</th>\n",
       "      <th>description</th>\n",
       "      <th>ingredients</th>\n",
       "      <th>n_ingredients</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>amish  tomato ketchup  for canning</td>\n",
       "      <td>44061</td>\n",
       "      <td>190</td>\n",
       "      <td>41706</td>\n",
       "      <td>2002-10-25</td>\n",
       "      <td>['weeknight', 'time-to-make', 'course', 'main-...</td>\n",
       "      <td>[352.9, 1.0, 337.0, 23.0, 3.0, 0.0, 28.0]</td>\n",
       "      <td>5</td>\n",
       "      <td>['mix all ingredients&amp; boil for 2 1 / 2 hours ...</td>\n",
       "      <td>my dh's amish mother raised him on this recipe...</td>\n",
       "      <td>['tomato juice', 'apple cider vinegar', 'sugar...</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>all in the kitchen  chili</td>\n",
       "      <td>112140</td>\n",
       "      <td>130</td>\n",
       "      <td>196586</td>\n",
       "      <td>2005-02-25</td>\n",
       "      <td>['time-to-make', 'course', 'preparation', 'mai...</td>\n",
       "      <td>[269.8, 22.0, 32.0, 48.0, 39.0, 27.0, 5.0]</td>\n",
       "      <td>6</td>\n",
       "      <td>['brown ground beef in large pot', 'add choppe...</td>\n",
       "      <td>this modified version of 'mom's' chili was a h...</td>\n",
       "      <td>['ground beef', 'yellow onions', 'diced tomato...</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>a bit different  breakfast pizza</td>\n",
       "      <td>31490</td>\n",
       "      <td>30</td>\n",
       "      <td>26278</td>\n",
       "      <td>2002-06-17</td>\n",
       "      <td>['30-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>[173.4, 18.0, 0.0, 17.0, 22.0, 35.0, 1.0]</td>\n",
       "      <td>9</td>\n",
       "      <td>['preheat oven to 425 degrees f', 'press dough...</td>\n",
       "      <td>this recipe calls for the crust to be prebaked...</td>\n",
       "      <td>['prepared pizza crust', 'sausage patty', 'egg...</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>arriba   baked winter squash mexican style</td>\n",
       "      <td>137739</td>\n",
       "      <td>55</td>\n",
       "      <td>47892</td>\n",
       "      <td>2005-09-16</td>\n",
       "      <td>['60-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>[51.5, 0.0, 13.0, 0.0, 2.0, 0.0, 4.0]</td>\n",
       "      <td>11</td>\n",
       "      <td>['make a choice and proceed with recipe', 'dep...</td>\n",
       "      <td>autumn is my favorite time of year to cook! th...</td>\n",
       "      <td>['winter squash', 'mexican seasoning', 'mixed ...</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>alouette  potatoes</td>\n",
       "      <td>59389</td>\n",
       "      <td>45</td>\n",
       "      <td>68585</td>\n",
       "      <td>2003-04-14</td>\n",
       "      <td>['60-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>[368.1, 17.0, 10.0, 2.0, 14.0, 8.0, 20.0]</td>\n",
       "      <td>11</td>\n",
       "      <td>['place potatoes in a large pot of lightly sal...</td>\n",
       "      <td>this is a super easy, great tasting, make ahea...</td>\n",
       "      <td>['spreadable cheese with garlic and herbs', 'n...</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         name      id  minutes contributor_id  \\\n",
       "4          amish  tomato ketchup  for canning   44061      190          41706   \n",
       "2                   all in the kitchen  chili  112140      130         196586   \n",
       "1            a bit different  breakfast pizza   31490       30          26278   \n",
       "0  arriba   baked winter squash mexican style  137739       55          47892   \n",
       "3                          alouette  potatoes   59389       45          68585   \n",
       "\n",
       "   submitted                                               tags  \\\n",
       "4 2002-10-25  ['weeknight', 'time-to-make', 'course', 'main-...   \n",
       "2 2005-02-25  ['time-to-make', 'course', 'preparation', 'mai...   \n",
       "1 2002-06-17  ['30-minutes-or-less', 'time-to-make', 'course...   \n",
       "0 2005-09-16  ['60-minutes-or-less', 'time-to-make', 'course...   \n",
       "3 2003-04-14  ['60-minutes-or-less', 'time-to-make', 'course...   \n",
       "\n",
       "                                    nutrition  n_steps  \\\n",
       "4   [352.9, 1.0, 337.0, 23.0, 3.0, 0.0, 28.0]        5   \n",
       "2  [269.8, 22.0, 32.0, 48.0, 39.0, 27.0, 5.0]        6   \n",
       "1   [173.4, 18.0, 0.0, 17.0, 22.0, 35.0, 1.0]        9   \n",
       "0       [51.5, 0.0, 13.0, 0.0, 2.0, 0.0, 4.0]       11   \n",
       "3   [368.1, 17.0, 10.0, 2.0, 14.0, 8.0, 20.0]       11   \n",
       "\n",
       "                                               steps  \\\n",
       "4  ['mix all ingredients& boil for 2 1 / 2 hours ...   \n",
       "2  ['brown ground beef in large pot', 'add choppe...   \n",
       "1  ['preheat oven to 425 degrees f', 'press dough...   \n",
       "0  ['make a choice and proceed with recipe', 'dep...   \n",
       "3  ['place potatoes in a large pot of lightly sal...   \n",
       "\n",
       "                                         description  \\\n",
       "4  my dh's amish mother raised him on this recipe...   \n",
       "2  this modified version of 'mom's' chili was a h...   \n",
       "1  this recipe calls for the crust to be prebaked...   \n",
       "0  autumn is my favorite time of year to cook! th...   \n",
       "3  this is a super easy, great tasting, make ahea...   \n",
       "\n",
       "                                         ingredients  n_ingredients  \n",
       "4  ['tomato juice', 'apple cider vinegar', 'sugar...              8  \n",
       "2  ['ground beef', 'yellow onions', 'diced tomato...             13  \n",
       "1  ['prepared pizza crust', 'sausage patty', 'egg...              6  \n",
       "0  ['winter squash', 'mexican seasoning', 'mixed ...              7  \n",
       "3  ['spreadable cheese with garlic and herbs', 'n...             11  "
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check dataframe\n",
    "raw_recipes_df.head().sort_values('n_steps')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d4403155",
   "metadata": {},
   "source": [
    "We will proceed to keep only the recipes with at least 1 step."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "4b7c9e97",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Filter for recipes with at least 1 step\n",
    "raw_recipes_df = raw_recipes_df[raw_recipes_df['n_steps'] > 0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "ac232a3e",
   "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>name</th>\n",
       "      <th>id</th>\n",
       "      <th>minutes</th>\n",
       "      <th>contributor_id</th>\n",
       "      <th>submitted</th>\n",
       "      <th>tags</th>\n",
       "      <th>nutrition</th>\n",
       "      <th>n_steps</th>\n",
       "      <th>steps</th>\n",
       "      <th>description</th>\n",
       "      <th>ingredients</th>\n",
       "      <th>n_ingredients</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>amish  tomato ketchup  for canning</td>\n",
       "      <td>44061</td>\n",
       "      <td>190</td>\n",
       "      <td>41706</td>\n",
       "      <td>2002-10-25</td>\n",
       "      <td>['weeknight', 'time-to-make', 'course', 'main-...</td>\n",
       "      <td>[352.9, 1.0, 337.0, 23.0, 3.0, 0.0, 28.0]</td>\n",
       "      <td>5</td>\n",
       "      <td>['mix all ingredients&amp; boil for 2 1 / 2 hours ...</td>\n",
       "      <td>my dh's amish mother raised him on this recipe...</td>\n",
       "      <td>['tomato juice', 'apple cider vinegar', 'sugar...</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>all in the kitchen  chili</td>\n",
       "      <td>112140</td>\n",
       "      <td>130</td>\n",
       "      <td>196586</td>\n",
       "      <td>2005-02-25</td>\n",
       "      <td>['time-to-make', 'course', 'preparation', 'mai...</td>\n",
       "      <td>[269.8, 22.0, 32.0, 48.0, 39.0, 27.0, 5.0]</td>\n",
       "      <td>6</td>\n",
       "      <td>['brown ground beef in large pot', 'add choppe...</td>\n",
       "      <td>this modified version of 'mom's' chili was a h...</td>\n",
       "      <td>['ground beef', 'yellow onions', 'diced tomato...</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>a bit different  breakfast pizza</td>\n",
       "      <td>31490</td>\n",
       "      <td>30</td>\n",
       "      <td>26278</td>\n",
       "      <td>2002-06-17</td>\n",
       "      <td>['30-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>[173.4, 18.0, 0.0, 17.0, 22.0, 35.0, 1.0]</td>\n",
       "      <td>9</td>\n",
       "      <td>['preheat oven to 425 degrees f', 'press dough...</td>\n",
       "      <td>this recipe calls for the crust to be prebaked...</td>\n",
       "      <td>['prepared pizza crust', 'sausage patty', 'egg...</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>arriba   baked winter squash mexican style</td>\n",
       "      <td>137739</td>\n",
       "      <td>55</td>\n",
       "      <td>47892</td>\n",
       "      <td>2005-09-16</td>\n",
       "      <td>['60-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>[51.5, 0.0, 13.0, 0.0, 2.0, 0.0, 4.0]</td>\n",
       "      <td>11</td>\n",
       "      <td>['make a choice and proceed with recipe', 'dep...</td>\n",
       "      <td>autumn is my favorite time of year to cook! th...</td>\n",
       "      <td>['winter squash', 'mexican seasoning', 'mixed ...</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>alouette  potatoes</td>\n",
       "      <td>59389</td>\n",
       "      <td>45</td>\n",
       "      <td>68585</td>\n",
       "      <td>2003-04-14</td>\n",
       "      <td>['60-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>[368.1, 17.0, 10.0, 2.0, 14.0, 8.0, 20.0]</td>\n",
       "      <td>11</td>\n",
       "      <td>['place potatoes in a large pot of lightly sal...</td>\n",
       "      <td>this is a super easy, great tasting, make ahea...</td>\n",
       "      <td>['spreadable cheese with garlic and herbs', 'n...</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         name      id  minutes contributor_id  \\\n",
       "4          amish  tomato ketchup  for canning   44061      190          41706   \n",
       "2                   all in the kitchen  chili  112140      130         196586   \n",
       "1            a bit different  breakfast pizza   31490       30          26278   \n",
       "0  arriba   baked winter squash mexican style  137739       55          47892   \n",
       "3                          alouette  potatoes   59389       45          68585   \n",
       "\n",
       "   submitted                                               tags  \\\n",
       "4 2002-10-25  ['weeknight', 'time-to-make', 'course', 'main-...   \n",
       "2 2005-02-25  ['time-to-make', 'course', 'preparation', 'mai...   \n",
       "1 2002-06-17  ['30-minutes-or-less', 'time-to-make', 'course...   \n",
       "0 2005-09-16  ['60-minutes-or-less', 'time-to-make', 'course...   \n",
       "3 2003-04-14  ['60-minutes-or-less', 'time-to-make', 'course...   \n",
       "\n",
       "                                    nutrition  n_steps  \\\n",
       "4   [352.9, 1.0, 337.0, 23.0, 3.0, 0.0, 28.0]        5   \n",
       "2  [269.8, 22.0, 32.0, 48.0, 39.0, 27.0, 5.0]        6   \n",
       "1   [173.4, 18.0, 0.0, 17.0, 22.0, 35.0, 1.0]        9   \n",
       "0       [51.5, 0.0, 13.0, 0.0, 2.0, 0.0, 4.0]       11   \n",
       "3   [368.1, 17.0, 10.0, 2.0, 14.0, 8.0, 20.0]       11   \n",
       "\n",
       "                                               steps  \\\n",
       "4  ['mix all ingredients& boil for 2 1 / 2 hours ...   \n",
       "2  ['brown ground beef in large pot', 'add choppe...   \n",
       "1  ['preheat oven to 425 degrees f', 'press dough...   \n",
       "0  ['make a choice and proceed with recipe', 'dep...   \n",
       "3  ['place potatoes in a large pot of lightly sal...   \n",
       "\n",
       "                                         description  \\\n",
       "4  my dh's amish mother raised him on this recipe...   \n",
       "2  this modified version of 'mom's' chili was a h...   \n",
       "1  this recipe calls for the crust to be prebaked...   \n",
       "0  autumn is my favorite time of year to cook! th...   \n",
       "3  this is a super easy, great tasting, make ahea...   \n",
       "\n",
       "                                         ingredients  n_ingredients  \n",
       "4  ['tomato juice', 'apple cider vinegar', 'sugar...              8  \n",
       "2  ['ground beef', 'yellow onions', 'diced tomato...             13  \n",
       "1  ['prepared pizza crust', 'sausage patty', 'egg...              6  \n",
       "0  ['winter squash', 'mexican seasoning', 'mixed ...              7  \n",
       "3  ['spreadable cheese with garlic and herbs', 'n...             11  "
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Sanity check\n",
    "raw_recipes_df.head().sort_values('n_steps')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e7e18400",
   "metadata": {},
   "source": [
    "Let us now look at the number of ingredients."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "f04886cd",
   "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>name</th>\n",
       "      <th>id</th>\n",
       "      <th>minutes</th>\n",
       "      <th>contributor_id</th>\n",
       "      <th>submitted</th>\n",
       "      <th>tags</th>\n",
       "      <th>nutrition</th>\n",
       "      <th>n_steps</th>\n",
       "      <th>steps</th>\n",
       "      <th>description</th>\n",
       "      <th>ingredients</th>\n",
       "      <th>n_ingredients</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>95755</th>\n",
       "      <td>granola extreme</td>\n",
       "      <td>43916</td>\n",
       "      <td>45</td>\n",
       "      <td>27443</td>\n",
       "      <td>2002-10-22</td>\n",
       "      <td>['60-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>[105.0, 10.0, 10.0, 2.0, 5.0, 5.0, 3.0]</td>\n",
       "      <td>9</td>\n",
       "      <td>[\"preheat oven 300' degrees\", 'spray 2 baking ...</td>\n",
       "      <td>wholesome snack, great topping for vanilla yogurt</td>\n",
       "      <td>['oatmeal', 'wheat flakes', 'buckwheat flakes'...</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>137311</th>\n",
       "      <td>mocha mini bread loaves</td>\n",
       "      <td>349861</td>\n",
       "      <td>40</td>\n",
       "      <td>1109972</td>\n",
       "      <td>2009-01-14</td>\n",
       "      <td>['60-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>[268.4, 12.0, 75.0, 12.0, 14.0, 11.0, 14.0]</td>\n",
       "      <td>7</td>\n",
       "      <td>['preheat oven to 350 degrees', 'combine dry i...</td>\n",
       "      <td>this is adapted from camilla saulsbury's muffi...</td>\n",
       "      <td>['whole wheat flour', 'quick-cooking oats', 't...</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68576</th>\n",
       "      <td>curried meatballs</td>\n",
       "      <td>315007</td>\n",
       "      <td>35</td>\n",
       "      <td>331557</td>\n",
       "      <td>2008-07-22</td>\n",
       "      <td>['curries', '60-minutes-or-less', 'time-to-mak...</td>\n",
       "      <td>[414.4, 36.0, 21.0, 32.0, 59.0, 34.0, 6.0]</td>\n",
       "      <td>8</td>\n",
       "      <td>['heat oil over medium heat and stir-fry the o...</td>\n",
       "      <td>i was given this recipe by a co-worker of mine...</td>\n",
       "      <td>['olive oil', 'onions', 'cinnamon stick', 'wat...</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>214382</th>\n",
       "      <td>tofu spinach calzones  vegan or vegetarian</td>\n",
       "      <td>97154</td>\n",
       "      <td>35</td>\n",
       "      <td>37584</td>\n",
       "      <td>2004-08-06</td>\n",
       "      <td>['60-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>[346.1, 6.0, 8.0, 14.0, 21.0, 3.0, 21.0]</td>\n",
       "      <td>17</td>\n",
       "      <td>['mix water with yeast and sugar in a large mi...</td>\n",
       "      <td>these are great and vegan or vegetarian (see s...</td>\n",
       "      <td>['warm water', 'yeast', 'sugar', 'salt', 'flou...</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68671</th>\n",
       "      <td>curried rice and fruit salad</td>\n",
       "      <td>63805</td>\n",
       "      <td>65</td>\n",
       "      <td>37779</td>\n",
       "      <td>2003-06-05</td>\n",
       "      <td>['weeknight', 'time-to-make', 'course', 'main-...</td>\n",
       "      <td>[379.1, 11.0, 86.0, 16.0, 12.0, 14.0, 24.0]</td>\n",
       "      <td>18</td>\n",
       "      <td>['wash rice until water runs clear', 'drain we...</td>\n",
       "      <td>this rice salad has exotic flavors that make a...</td>\n",
       "      <td>['basmati rice', 'peanut oil', 'green onion', ...</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>105723</th>\n",
       "      <td>hillbilly bean soup</td>\n",
       "      <td>11737</td>\n",
       "      <td>255</td>\n",
       "      <td>10404</td>\n",
       "      <td>2001-09-18</td>\n",
       "      <td>['weeknight', 'time-to-make', 'course', 'main-...</td>\n",
       "      <td>[323.0, 2.0, 254.0, 1.0, 8.0, 1.0, 27.0]</td>\n",
       "      <td>6</td>\n",
       "      <td>['boil the drained , soaked beans with the wat...</td>\n",
       "      <td>great soup for a fall or winters dinner. serve...</td>\n",
       "      <td>['7 bean mix', 'water', 'ham bone', 'bay leaf'...</td>\n",
       "      <td>35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>203614</th>\n",
       "      <td>sunday supper taco salad bowls   tacos</td>\n",
       "      <td>502219</td>\n",
       "      <td>240</td>\n",
       "      <td>695000</td>\n",
       "      <td>2013-06-23</td>\n",
       "      <td>['time-to-make', 'course', 'cuisine', 'prepara...</td>\n",
       "      <td>[10687.7, 1784.0, 49.0, 67.0, 59.0, 802.0, 26.0]</td>\n",
       "      <td>6</td>\n",
       "      <td>['heat vegetable oil to 350f submerge tortilla...</td>\n",
       "      <td>being served again at the free sunday supper o...</td>\n",
       "      <td>['flour tortillas', 'ground beef', 'vegetable ...</td>\n",
       "      <td>37</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18352</th>\n",
       "      <td>basic fried rice   with variations</td>\n",
       "      <td>248707</td>\n",
       "      <td>55</td>\n",
       "      <td>305531</td>\n",
       "      <td>2007-08-25</td>\n",
       "      <td>['60-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>[793.3, 22.0, 8.0, 23.0, 37.0, 13.0, 48.0]</td>\n",
       "      <td>16</td>\n",
       "      <td>['dice or shred meat and vegetables', 'heat oi...</td>\n",
       "      <td>another recipe that i acquired from a chinese ...</td>\n",
       "      <td>['brown rice', 'green onions', 'oil', 'soy sau...</td>\n",
       "      <td>38</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>116744</th>\n",
       "      <td>kato s  garage  style    tomato sauce   meatba...</td>\n",
       "      <td>113677</td>\n",
       "      <td>270</td>\n",
       "      <td>128473</td>\n",
       "      <td>2005-03-18</td>\n",
       "      <td>['time-to-make', 'course', 'main-ingredient', ...</td>\n",
       "      <td>[412.8, 36.0, 61.0, 47.0, 46.0, 38.0, 8.0]</td>\n",
       "      <td>9</td>\n",
       "      <td>['spicy tomatoe &amp; vegetable sauce', 'in a larg...</td>\n",
       "      <td>this is our favourite spagetti dinner...we mak...</td>\n",
       "      <td>['sweet onion', 'garlic cloves', 'green pepper...</td>\n",
       "      <td>39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2162</th>\n",
       "      <td>a to z bread</td>\n",
       "      <td>3748</td>\n",
       "      <td>80</td>\n",
       "      <td>1535</td>\n",
       "      <td>1999-10-30</td>\n",
       "      <td>['time-to-make', 'course', 'main-ingredient', ...</td>\n",
       "      <td>[367.0, 30.0, 101.0, 10.0, 9.0, 12.0, 14.0]</td>\n",
       "      <td>12</td>\n",
       "      <td>['in a mixing bowl , combine the sugar , oil a...</td>\n",
       "      <td>an extremely versatile bread. very moist and d...</td>\n",
       "      <td>['sugar', 'vegetable oil', 'eggs', 'vanilla ex...</td>\n",
       "      <td>43</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2685 rows × 12 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                     name      id  minutes  \\\n",
       "95755                                     granola extreme   43916       45   \n",
       "137311                            mocha mini bread loaves  349861       40   \n",
       "68576                                   curried meatballs  315007       35   \n",
       "214382         tofu spinach calzones  vegan or vegetarian   97154       35   \n",
       "68671                        curried rice and fruit salad   63805       65   \n",
       "...                                                   ...     ...      ...   \n",
       "105723                                hillbilly bean soup   11737      255   \n",
       "203614             sunday supper taco salad bowls   tacos  502219      240   \n",
       "18352                  basic fried rice   with variations  248707       55   \n",
       "116744  kato s  garage  style    tomato sauce   meatba...  113677      270   \n",
       "2162                                         a to z bread    3748       80   \n",
       "\n",
       "       contributor_id  submitted  \\\n",
       "95755           27443 2002-10-22   \n",
       "137311        1109972 2009-01-14   \n",
       "68576          331557 2008-07-22   \n",
       "214382          37584 2004-08-06   \n",
       "68671           37779 2003-06-05   \n",
       "...               ...        ...   \n",
       "105723          10404 2001-09-18   \n",
       "203614         695000 2013-06-23   \n",
       "18352          305531 2007-08-25   \n",
       "116744         128473 2005-03-18   \n",
       "2162             1535 1999-10-30   \n",
       "\n",
       "                                                     tags  \\\n",
       "95755   ['60-minutes-or-less', 'time-to-make', 'course...   \n",
       "137311  ['60-minutes-or-less', 'time-to-make', 'course...   \n",
       "68576   ['curries', '60-minutes-or-less', 'time-to-mak...   \n",
       "214382  ['60-minutes-or-less', 'time-to-make', 'course...   \n",
       "68671   ['weeknight', 'time-to-make', 'course', 'main-...   \n",
       "...                                                   ...   \n",
       "105723  ['weeknight', 'time-to-make', 'course', 'main-...   \n",
       "203614  ['time-to-make', 'course', 'cuisine', 'prepara...   \n",
       "18352   ['60-minutes-or-less', 'time-to-make', 'course...   \n",
       "116744  ['time-to-make', 'course', 'main-ingredient', ...   \n",
       "2162    ['time-to-make', 'course', 'main-ingredient', ...   \n",
       "\n",
       "                                               nutrition  n_steps  \\\n",
       "95755            [105.0, 10.0, 10.0, 2.0, 5.0, 5.0, 3.0]        9   \n",
       "137311       [268.4, 12.0, 75.0, 12.0, 14.0, 11.0, 14.0]        7   \n",
       "68576         [414.4, 36.0, 21.0, 32.0, 59.0, 34.0, 6.0]        8   \n",
       "214382          [346.1, 6.0, 8.0, 14.0, 21.0, 3.0, 21.0]       17   \n",
       "68671        [379.1, 11.0, 86.0, 16.0, 12.0, 14.0, 24.0]       18   \n",
       "...                                                  ...      ...   \n",
       "105723          [323.0, 2.0, 254.0, 1.0, 8.0, 1.0, 27.0]        6   \n",
       "203614  [10687.7, 1784.0, 49.0, 67.0, 59.0, 802.0, 26.0]        6   \n",
       "18352         [793.3, 22.0, 8.0, 23.0, 37.0, 13.0, 48.0]       16   \n",
       "116744        [412.8, 36.0, 61.0, 47.0, 46.0, 38.0, 8.0]        9   \n",
       "2162         [367.0, 30.0, 101.0, 10.0, 9.0, 12.0, 14.0]       12   \n",
       "\n",
       "                                                    steps  \\\n",
       "95755   [\"preheat oven 300' degrees\", 'spray 2 baking ...   \n",
       "137311  ['preheat oven to 350 degrees', 'combine dry i...   \n",
       "68576   ['heat oil over medium heat and stir-fry the o...   \n",
       "214382  ['mix water with yeast and sugar in a large mi...   \n",
       "68671   ['wash rice until water runs clear', 'drain we...   \n",
       "...                                                   ...   \n",
       "105723  ['boil the drained , soaked beans with the wat...   \n",
       "203614  ['heat vegetable oil to 350f submerge tortilla...   \n",
       "18352   ['dice or shred meat and vegetables', 'heat oi...   \n",
       "116744  ['spicy tomatoe & vegetable sauce', 'in a larg...   \n",
       "2162    ['in a mixing bowl , combine the sugar , oil a...   \n",
       "\n",
       "                                              description  \\\n",
       "95755   wholesome snack, great topping for vanilla yogurt   \n",
       "137311  this is adapted from camilla saulsbury's muffi...   \n",
       "68576   i was given this recipe by a co-worker of mine...   \n",
       "214382  these are great and vegan or vegetarian (see s...   \n",
       "68671   this rice salad has exotic flavors that make a...   \n",
       "...                                                   ...   \n",
       "105723  great soup for a fall or winters dinner. serve...   \n",
       "203614  being served again at the free sunday supper o...   \n",
       "18352   another recipe that i acquired from a chinese ...   \n",
       "116744  this is our favourite spagetti dinner...we mak...   \n",
       "2162    an extremely versatile bread. very moist and d...   \n",
       "\n",
       "                                              ingredients  n_ingredients  \n",
       "95755   ['oatmeal', 'wheat flakes', 'buckwheat flakes'...             19  \n",
       "137311  ['whole wheat flour', 'quick-cooking oats', 't...             19  \n",
       "68576   ['olive oil', 'onions', 'cinnamon stick', 'wat...             19  \n",
       "214382  ['warm water', 'yeast', 'sugar', 'salt', 'flou...             19  \n",
       "68671   ['basmati rice', 'peanut oil', 'green onion', ...             19  \n",
       "...                                                   ...            ...  \n",
       "105723  ['7 bean mix', 'water', 'ham bone', 'bay leaf'...             35  \n",
       "203614  ['flour tortillas', 'ground beef', 'vegetable ...             37  \n",
       "18352   ['brown rice', 'green onions', 'oil', 'soy sau...             38  \n",
       "116744  ['sweet onion', 'garlic cloves', 'green pepper...             39  \n",
       "2162    ['sugar', 'vegetable oil', 'eggs', 'vanilla ex...             43  \n",
       "\n",
       "[2685 rows x 12 columns]"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Number of ingredients\n",
    "# Calculate the first quartile, third quartile and the interquartile range\n",
    "Q1 = raw_recipes_df['n_ingredients'].quantile(0.25)\n",
    "Q3 = raw_recipes_df['n_ingredients'].quantile(0.75)\n",
    "IQR = Q3 - Q1\n",
    "\n",
    "# Calculate the maximum value and minimum values according to the Tukey rule\n",
    "max_value = Q3 + 1.5 * IQR\n",
    "min_value = Q1 - 1.5 * IQR\n",
    "\n",
    "# Filter the data for values that are greater than max_value or less than min_value\n",
    "nsteps_outliers = raw_recipes_df[(raw_recipes_df['n_ingredients'] > max_value) | (raw_recipes_df['n_ingredients'] < min_value)]\n",
    "nsteps_outliers.sort_values('n_ingredients')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "f04662c5",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualize boxen plot\n",
    "sns.set(style = \"whitegrid\")\n",
    "ax = sns.boxenplot(x = raw_recipes_df['n_ingredients'])\n",
    "ax.set_xlabel('Number of steps per recipe')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4c1936eb",
   "metadata": {},
   "source": [
    "A to Z Bread holds the distinction for having the highest number of ingredients among all recipes in the dataset. However, it is worth noting that a significant portion of these ingredients are labeled as optional. Although it deviates significantly from the average number of ingredients in the recipes, it is important to retain this data point to maintain the integrity and representativeness of the dataset. The presence of recipes with a higher number of optional ingredients adds variety and reflects the diverse nature of cooking and recipe preferences. By retaining this outlier, we ensure that our analysis encompasses a wide range of recipes and provides a comprehensive understanding of the dataset."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2b8f23ae",
   "metadata": {},
   "source": [
    "Let us now explore the user interactions dataset to gain insights into the ratings."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "6ad5a6b3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Distribution of ratings\n",
    "\n",
    "rating_counts = raw_users_df['rating'].value_counts().sort_index()\n",
    "ratings = rating_counts.index\n",
    "counts = rating_counts.values\n",
    "\n",
    "plt.bar(ratings, counts)\n",
    "plt.xlabel('Rating')\n",
    "plt.ylabel('Count')\n",
    "plt.title('Distribution of Ratings in User Interactions')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8df75940",
   "metadata": {},
   "source": [
    "We see that there are zero ratings. In the upcoming section, we will delve deeper into this aspect to gain a better understanding of the reasons behind these zero ratings and its implications."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1c42867c",
   "metadata": {},
   "source": [
    "### 3.5 Zero Ratings <a class=\"anchor\" id=\"threefive\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "95cc7ecf",
   "metadata": {},
   "source": [
    "After further investigation, we have discovered from the source that zero ratings in the user interactions dataset represent reviews by users who have not yet tried the respective recipes. As our focus is on reviews from users who have actually tested the recipes, we will proceed to remove these zero-rated reviews from our dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "1850bc98",
   "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>recipe_id</th>\n",
       "      <th>date</th>\n",
       "      <th>rating</th>\n",
       "      <th>review</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>704658</th>\n",
       "      <td>1762248</td>\n",
       "      <td>428870</td>\n",
       "      <td>2013-09-10</td>\n",
       "      <td>1</td>\n",
       "      <td>These were so awful that my husband threw them...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>305820</th>\n",
       "      <td>856238</td>\n",
       "      <td>213319</td>\n",
       "      <td>2008-06-07</td>\n",
       "      <td>1</td>\n",
       "      <td>There are no stars for less than \"one\".  Not g...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>305823</th>\n",
       "      <td>1263812</td>\n",
       "      <td>213319</td>\n",
       "      <td>2009-07-02</td>\n",
       "      <td>1</td>\n",
       "      <td>Yuck!  While it's palatable, it's not fudge-li...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>305824</th>\n",
       "      <td>538920</td>\n",
       "      <td>213319</td>\n",
       "      <td>2009-08-07</td>\n",
       "      <td>1</td>\n",
       "      <td>Some things just should simply not be made low...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>822728</th>\n",
       "      <td>4395</td>\n",
       "      <td>4829</td>\n",
       "      <td>2001-01-10</td>\n",
       "      <td>1</td>\n",
       "      <td>I believe there is a mistake in the amount of ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>428408</th>\n",
       "      <td>431813</td>\n",
       "      <td>171157</td>\n",
       "      <td>2007-06-27</td>\n",
       "      <td>5</td>\n",
       "      <td>Wow!  This far exceeded my expectations.  The ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>428409</th>\n",
       "      <td>546894</td>\n",
       "      <td>171157</td>\n",
       "      <td>2007-07-27</td>\n",
       "      <td>5</td>\n",
       "      <td>I have made this several times for guests and ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>428410</th>\n",
       "      <td>819967</td>\n",
       "      <td>171157</td>\n",
       "      <td>2009-05-19</td>\n",
       "      <td>5</td>\n",
       "      <td>Amazing recipe, this will be my standard recip...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>428365</th>\n",
       "      <td>110433</td>\n",
       "      <td>22149</td>\n",
       "      <td>2007-05-13</td>\n",
       "      <td>5</td>\n",
       "      <td>This was so juicy and had just a little kick t...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1132366</th>\n",
       "      <td>2001868099</td>\n",
       "      <td>78003</td>\n",
       "      <td>2017-12-18</td>\n",
       "      <td>5</td>\n",
       "      <td>I am so glad I googled and found this here. Th...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1071351 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            user_id recipe_id       date  rating  \\\n",
       "704658      1762248    428870 2013-09-10       1   \n",
       "305820       856238    213319 2008-06-07       1   \n",
       "305823      1263812    213319 2009-07-02       1   \n",
       "305824       538920    213319 2009-08-07       1   \n",
       "822728         4395      4829 2001-01-10       1   \n",
       "...             ...       ...        ...     ...   \n",
       "428408       431813    171157 2007-06-27       5   \n",
       "428409       546894    171157 2007-07-27       5   \n",
       "428410       819967    171157 2009-05-19       5   \n",
       "428365       110433     22149 2007-05-13       5   \n",
       "1132366  2001868099     78003 2017-12-18       5   \n",
       "\n",
       "                                                    review  \n",
       "704658   These were so awful that my husband threw them...  \n",
       "305820   There are no stars for less than \"one\".  Not g...  \n",
       "305823   Yuck!  While it's palatable, it's not fudge-li...  \n",
       "305824   Some things just should simply not be made low...  \n",
       "822728   I believe there is a mistake in the amount of ...  \n",
       "...                                                    ...  \n",
       "428408   Wow!  This far exceeded my expectations.  The ...  \n",
       "428409   I have made this several times for guests and ...  \n",
       "428410   Amazing recipe, this will be my standard recip...  \n",
       "428365   This was so juicy and had just a little kick t...  \n",
       "1132366  I am so glad I googled and found this here. Th...  \n",
       "\n",
       "[1071351 rows x 5 columns]"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Keep rows that have ratings between 1 and 5\n",
    "raw_users_df = raw_users_df[raw_users_df['rating'] != 0]\n",
    "raw_users_df.sort_values('rating')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0e7d1cf2",
   "metadata": {},
   "source": [
    "## 4 Data Preprocessing <a class=\"anchor\" id=\"four\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ab767f64",
   "metadata": {},
   "source": [
    "After completing the data cleaning phase, we will proceed to the data preprocessing stage. Let's revisit the recipes dataframe to review its current state."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "d2be6f1c",
   "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>name</th>\n",
       "      <th>id</th>\n",
       "      <th>minutes</th>\n",
       "      <th>contributor_id</th>\n",
       "      <th>submitted</th>\n",
       "      <th>tags</th>\n",
       "      <th>nutrition</th>\n",
       "      <th>n_steps</th>\n",
       "      <th>steps</th>\n",
       "      <th>description</th>\n",
       "      <th>ingredients</th>\n",
       "      <th>n_ingredients</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>arriba   baked winter squash mexican style</td>\n",
       "      <td>137739</td>\n",
       "      <td>55</td>\n",
       "      <td>47892</td>\n",
       "      <td>2005-09-16</td>\n",
       "      <td>['60-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>[51.5, 0.0, 13.0, 0.0, 2.0, 0.0, 4.0]</td>\n",
       "      <td>11</td>\n",
       "      <td>['make a choice and proceed with recipe', 'dep...</td>\n",
       "      <td>autumn is my favorite time of year to cook! th...</td>\n",
       "      <td>['winter squash', 'mexican seasoning', 'mixed ...</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>a bit different  breakfast pizza</td>\n",
       "      <td>31490</td>\n",
       "      <td>30</td>\n",
       "      <td>26278</td>\n",
       "      <td>2002-06-17</td>\n",
       "      <td>['30-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>[173.4, 18.0, 0.0, 17.0, 22.0, 35.0, 1.0]</td>\n",
       "      <td>9</td>\n",
       "      <td>['preheat oven to 425 degrees f', 'press dough...</td>\n",
       "      <td>this recipe calls for the crust to be prebaked...</td>\n",
       "      <td>['prepared pizza crust', 'sausage patty', 'egg...</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>all in the kitchen  chili</td>\n",
       "      <td>112140</td>\n",
       "      <td>130</td>\n",
       "      <td>196586</td>\n",
       "      <td>2005-02-25</td>\n",
       "      <td>['time-to-make', 'course', 'preparation', 'mai...</td>\n",
       "      <td>[269.8, 22.0, 32.0, 48.0, 39.0, 27.0, 5.0]</td>\n",
       "      <td>6</td>\n",
       "      <td>['brown ground beef in large pot', 'add choppe...</td>\n",
       "      <td>this modified version of 'mom's' chili was a h...</td>\n",
       "      <td>['ground beef', 'yellow onions', 'diced tomato...</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>alouette  potatoes</td>\n",
       "      <td>59389</td>\n",
       "      <td>45</td>\n",
       "      <td>68585</td>\n",
       "      <td>2003-04-14</td>\n",
       "      <td>['60-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>[368.1, 17.0, 10.0, 2.0, 14.0, 8.0, 20.0]</td>\n",
       "      <td>11</td>\n",
       "      <td>['place potatoes in a large pot of lightly sal...</td>\n",
       "      <td>this is a super easy, great tasting, make ahea...</td>\n",
       "      <td>['spreadable cheese with garlic and herbs', 'n...</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>amish  tomato ketchup  for canning</td>\n",
       "      <td>44061</td>\n",
       "      <td>190</td>\n",
       "      <td>41706</td>\n",
       "      <td>2002-10-25</td>\n",
       "      <td>['weeknight', 'time-to-make', 'course', 'main-...</td>\n",
       "      <td>[352.9, 1.0, 337.0, 23.0, 3.0, 0.0, 28.0]</td>\n",
       "      <td>5</td>\n",
       "      <td>['mix all ingredients&amp; boil for 2 1 / 2 hours ...</td>\n",
       "      <td>my dh's amish mother raised him on this recipe...</td>\n",
       "      <td>['tomato juice', 'apple cider vinegar', 'sugar...</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         name      id  minutes contributor_id  \\\n",
       "0  arriba   baked winter squash mexican style  137739       55          47892   \n",
       "1            a bit different  breakfast pizza   31490       30          26278   \n",
       "2                   all in the kitchen  chili  112140      130         196586   \n",
       "3                          alouette  potatoes   59389       45          68585   \n",
       "4          amish  tomato ketchup  for canning   44061      190          41706   \n",
       "\n",
       "   submitted                                               tags  \\\n",
       "0 2005-09-16  ['60-minutes-or-less', 'time-to-make', 'course...   \n",
       "1 2002-06-17  ['30-minutes-or-less', 'time-to-make', 'course...   \n",
       "2 2005-02-25  ['time-to-make', 'course', 'preparation', 'mai...   \n",
       "3 2003-04-14  ['60-minutes-or-less', 'time-to-make', 'course...   \n",
       "4 2002-10-25  ['weeknight', 'time-to-make', 'course', 'main-...   \n",
       "\n",
       "                                    nutrition  n_steps  \\\n",
       "0       [51.5, 0.0, 13.0, 0.0, 2.0, 0.0, 4.0]       11   \n",
       "1   [173.4, 18.0, 0.0, 17.0, 22.0, 35.0, 1.0]        9   \n",
       "2  [269.8, 22.0, 32.0, 48.0, 39.0, 27.0, 5.0]        6   \n",
       "3   [368.1, 17.0, 10.0, 2.0, 14.0, 8.0, 20.0]       11   \n",
       "4   [352.9, 1.0, 337.0, 23.0, 3.0, 0.0, 28.0]        5   \n",
       "\n",
       "                                               steps  \\\n",
       "0  ['make a choice and proceed with recipe', 'dep...   \n",
       "1  ['preheat oven to 425 degrees f', 'press dough...   \n",
       "2  ['brown ground beef in large pot', 'add choppe...   \n",
       "3  ['place potatoes in a large pot of lightly sal...   \n",
       "4  ['mix all ingredients& boil for 2 1 / 2 hours ...   \n",
       "\n",
       "                                         description  \\\n",
       "0  autumn is my favorite time of year to cook! th...   \n",
       "1  this recipe calls for the crust to be prebaked...   \n",
       "2  this modified version of 'mom's' chili was a h...   \n",
       "3  this is a super easy, great tasting, make ahea...   \n",
       "4  my dh's amish mother raised him on this recipe...   \n",
       "\n",
       "                                         ingredients  n_ingredients  \n",
       "0  ['winter squash', 'mexican seasoning', 'mixed ...              7  \n",
       "1  ['prepared pizza crust', 'sausage patty', 'egg...              6  \n",
       "2  ['ground beef', 'yellow onions', 'diced tomato...             13  \n",
       "3  ['spreadable cheese with garlic and herbs', 'n...             11  \n",
       "4  ['tomato juice', 'apple cider vinegar', 'sugar...              8  "
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# View recipes dataset\n",
    "raw_recipes_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9e3d9a1d",
   "metadata": {},
   "source": [
    "To enhance the dataset and gain more insights, we will perform a data transformation by splitting the values in the 'nutrition' column and creating new columns for each nutritional metric. Additionally, we will split the 'submitted' date into separate columns for month and year. These transformations will enable us to delve deeper into the nutritional aspects of the recipes and provide a more comprehensive analysis.\n",
    "\n",
    "Let us take a look at the user interactions dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "0d3c921b",
   "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>recipe_id</th>\n",
       "      <th>date</th>\n",
       "      <th>rating</th>\n",
       "      <th>review</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>38094</td>\n",
       "      <td>40893</td>\n",
       "      <td>2003-02-17</td>\n",
       "      <td>4</td>\n",
       "      <td>Great with a salad. Cooked on top of stove for...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1293707</td>\n",
       "      <td>40893</td>\n",
       "      <td>2011-12-21</td>\n",
       "      <td>5</td>\n",
       "      <td>So simple, so delicious! Great for chilly fall...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>8937</td>\n",
       "      <td>44394</td>\n",
       "      <td>2002-12-01</td>\n",
       "      <td>4</td>\n",
       "      <td>This worked very well and is EASY.  I used not...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>126440</td>\n",
       "      <td>85009</td>\n",
       "      <td>2010-02-27</td>\n",
       "      <td>5</td>\n",
       "      <td>I made the Mexican topping and took it to bunk...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>57222</td>\n",
       "      <td>85009</td>\n",
       "      <td>2011-10-01</td>\n",
       "      <td>5</td>\n",
       "      <td>Made the cheddar bacon topping, adding a sprin...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   user_id recipe_id       date  rating  \\\n",
       "0    38094     40893 2003-02-17       4   \n",
       "1  1293707     40893 2011-12-21       5   \n",
       "2     8937     44394 2002-12-01       4   \n",
       "3   126440     85009 2010-02-27       5   \n",
       "4    57222     85009 2011-10-01       5   \n",
       "\n",
       "                                              review  \n",
       "0  Great with a salad. Cooked on top of stove for...  \n",
       "1  So simple, so delicious! Great for chilly fall...  \n",
       "2  This worked very well and is EASY.  I used not...  \n",
       "3  I made the Mexican topping and took it to bunk...  \n",
       "4  Made the cheddar bacon topping, adding a sprin...  "
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# View user interactions dataset\n",
    "raw_users_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c7c7913b",
   "metadata": {},
   "source": [
    "The 'date' column in the user interactions dataset can also be separated into month and year columns. In the upcoming section, we will carry out these data transformations to further explore and understand the dataset."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f11b68dd",
   "metadata": {},
   "source": [
    "### 4.1 Feature Engineering <a class=\"anchor\" id=\"fourone\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f61ab450",
   "metadata": {},
   "source": [
    "#### 4.1.1 Nutritional Values <a class=\"anchor\" id=\"fouroneone\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3869508f",
   "metadata": {},
   "source": [
    "The 'nutrition' column in the dataframe consists of a list of values representing various nutritional metrics, including calories, total fat, sugar, sodium, protein, saturated fat, and carbohydrates. To facilitate further analysis, we have split these values and created new columns to capture each nutritional metric individually. Subsequently, we drop the original 'nutrition' column from the dataframe, as we have already extracted the nutritional metrics into separate columns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "e39bb85a",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Split the values in the 'nutrition' column into separate columns\n",
    "nutrition_columns = ['calories', 'total_fat', 'sugar', 'sodium', 'protein', 'saturated_fat', 'carbohydrates']\n",
    "\n",
    "for column in nutrition_columns:\n",
    "    raw_recipes_df[column] = raw_recipes_df['nutrition'].apply(lambda x: eval(x)[nutrition_columns.index(column)])\n",
    "\n",
    "# Drop the original 'nutrition' column\n",
    "raw_recipes_df.drop('nutrition', axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "6c5527e5",
   "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>name</th>\n",
       "      <th>id</th>\n",
       "      <th>minutes</th>\n",
       "      <th>contributor_id</th>\n",
       "      <th>submitted</th>\n",
       "      <th>tags</th>\n",
       "      <th>n_steps</th>\n",
       "      <th>steps</th>\n",
       "      <th>description</th>\n",
       "      <th>ingredients</th>\n",
       "      <th>n_ingredients</th>\n",
       "      <th>calories</th>\n",
       "      <th>total_fat</th>\n",
       "      <th>sugar</th>\n",
       "      <th>sodium</th>\n",
       "      <th>protein</th>\n",
       "      <th>saturated_fat</th>\n",
       "      <th>carbohydrates</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>arriba   baked winter squash mexican style</td>\n",
       "      <td>137739</td>\n",
       "      <td>55</td>\n",
       "      <td>47892</td>\n",
       "      <td>2005-09-16</td>\n",
       "      <td>['60-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>11</td>\n",
       "      <td>['make a choice and proceed with recipe', 'dep...</td>\n",
       "      <td>autumn is my favorite time of year to cook! th...</td>\n",
       "      <td>['winter squash', 'mexican seasoning', 'mixed ...</td>\n",
       "      <td>7</td>\n",
       "      <td>51.5</td>\n",
       "      <td>0.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>a bit different  breakfast pizza</td>\n",
       "      <td>31490</td>\n",
       "      <td>30</td>\n",
       "      <td>26278</td>\n",
       "      <td>2002-06-17</td>\n",
       "      <td>['30-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>9</td>\n",
       "      <td>['preheat oven to 425 degrees f', 'press dough...</td>\n",
       "      <td>this recipe calls for the crust to be prebaked...</td>\n",
       "      <td>['prepared pizza crust', 'sausage patty', 'egg...</td>\n",
       "      <td>6</td>\n",
       "      <td>173.4</td>\n",
       "      <td>18.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>22.0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>all in the kitchen  chili</td>\n",
       "      <td>112140</td>\n",
       "      <td>130</td>\n",
       "      <td>196586</td>\n",
       "      <td>2005-02-25</td>\n",
       "      <td>['time-to-make', 'course', 'preparation', 'mai...</td>\n",
       "      <td>6</td>\n",
       "      <td>['brown ground beef in large pot', 'add choppe...</td>\n",
       "      <td>this modified version of 'mom's' chili was a h...</td>\n",
       "      <td>['ground beef', 'yellow onions', 'diced tomato...</td>\n",
       "      <td>13</td>\n",
       "      <td>269.8</td>\n",
       "      <td>22.0</td>\n",
       "      <td>32.0</td>\n",
       "      <td>48.0</td>\n",
       "      <td>39.0</td>\n",
       "      <td>27.0</td>\n",
       "      <td>5.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>alouette  potatoes</td>\n",
       "      <td>59389</td>\n",
       "      <td>45</td>\n",
       "      <td>68585</td>\n",
       "      <td>2003-04-14</td>\n",
       "      <td>['60-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>11</td>\n",
       "      <td>['place potatoes in a large pot of lightly sal...</td>\n",
       "      <td>this is a super easy, great tasting, make ahea...</td>\n",
       "      <td>['spreadable cheese with garlic and herbs', 'n...</td>\n",
       "      <td>11</td>\n",
       "      <td>368.1</td>\n",
       "      <td>17.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>20.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>amish  tomato ketchup  for canning</td>\n",
       "      <td>44061</td>\n",
       "      <td>190</td>\n",
       "      <td>41706</td>\n",
       "      <td>2002-10-25</td>\n",
       "      <td>['weeknight', 'time-to-make', 'course', 'main-...</td>\n",
       "      <td>5</td>\n",
       "      <td>['mix all ingredients&amp; boil for 2 1 / 2 hours ...</td>\n",
       "      <td>my dh's amish mother raised him on this recipe...</td>\n",
       "      <td>['tomato juice', 'apple cider vinegar', 'sugar...</td>\n",
       "      <td>8</td>\n",
       "      <td>352.9</td>\n",
       "      <td>1.0</td>\n",
       "      <td>337.0</td>\n",
       "      <td>23.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>28.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         name      id  minutes contributor_id  \\\n",
       "0  arriba   baked winter squash mexican style  137739       55          47892   \n",
       "1            a bit different  breakfast pizza   31490       30          26278   \n",
       "2                   all in the kitchen  chili  112140      130         196586   \n",
       "3                          alouette  potatoes   59389       45          68585   \n",
       "4          amish  tomato ketchup  for canning   44061      190          41706   \n",
       "\n",
       "   submitted                                               tags  n_steps  \\\n",
       "0 2005-09-16  ['60-minutes-or-less', 'time-to-make', 'course...       11   \n",
       "1 2002-06-17  ['30-minutes-or-less', 'time-to-make', 'course...        9   \n",
       "2 2005-02-25  ['time-to-make', 'course', 'preparation', 'mai...        6   \n",
       "3 2003-04-14  ['60-minutes-or-less', 'time-to-make', 'course...       11   \n",
       "4 2002-10-25  ['weeknight', 'time-to-make', 'course', 'main-...        5   \n",
       "\n",
       "                                               steps  \\\n",
       "0  ['make a choice and proceed with recipe', 'dep...   \n",
       "1  ['preheat oven to 425 degrees f', 'press dough...   \n",
       "2  ['brown ground beef in large pot', 'add choppe...   \n",
       "3  ['place potatoes in a large pot of lightly sal...   \n",
       "4  ['mix all ingredients& boil for 2 1 / 2 hours ...   \n",
       "\n",
       "                                         description  \\\n",
       "0  autumn is my favorite time of year to cook! th...   \n",
       "1  this recipe calls for the crust to be prebaked...   \n",
       "2  this modified version of 'mom's' chili was a h...   \n",
       "3  this is a super easy, great tasting, make ahea...   \n",
       "4  my dh's amish mother raised him on this recipe...   \n",
       "\n",
       "                                         ingredients  n_ingredients  calories  \\\n",
       "0  ['winter squash', 'mexican seasoning', 'mixed ...              7      51.5   \n",
       "1  ['prepared pizza crust', 'sausage patty', 'egg...              6     173.4   \n",
       "2  ['ground beef', 'yellow onions', 'diced tomato...             13     269.8   \n",
       "3  ['spreadable cheese with garlic and herbs', 'n...             11     368.1   \n",
       "4  ['tomato juice', 'apple cider vinegar', 'sugar...              8     352.9   \n",
       "\n",
       "   total_fat  sugar  sodium  protein  saturated_fat  carbohydrates  \n",
       "0        0.0   13.0     0.0      2.0            0.0            4.0  \n",
       "1       18.0    0.0    17.0     22.0           35.0            1.0  \n",
       "2       22.0   32.0    48.0     39.0           27.0            5.0  \n",
       "3       17.0   10.0     2.0     14.0            8.0           20.0  \n",
       "4        1.0  337.0    23.0      3.0            0.0           28.0  "
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Sanity check\n",
    "raw_recipes_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "77bee0f4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 216818 entries, 0 to 231636\n",
      "Data columns (total 18 columns):\n",
      " #   Column          Non-Null Count   Dtype         \n",
      "---  ------          --------------   -----         \n",
      " 0   name            216818 non-null  object        \n",
      " 1   id              216818 non-null  object        \n",
      " 2   minutes         216818 non-null  int64         \n",
      " 3   contributor_id  216818 non-null  object        \n",
      " 4   submitted       216818 non-null  datetime64[ns]\n",
      " 5   tags            216818 non-null  object        \n",
      " 6   n_steps         216818 non-null  int64         \n",
      " 7   steps           216818 non-null  object        \n",
      " 8   description     216818 non-null  object        \n",
      " 9   ingredients     216818 non-null  object        \n",
      " 10  n_ingredients   216818 non-null  int64         \n",
      " 11  calories        216818 non-null  float64       \n",
      " 12  total_fat       216818 non-null  float64       \n",
      " 13  sugar           216818 non-null  float64       \n",
      " 14  sodium          216818 non-null  float64       \n",
      " 15  protein         216818 non-null  float64       \n",
      " 16  saturated_fat   216818 non-null  float64       \n",
      " 17  carbohydrates   216818 non-null  float64       \n",
      "dtypes: datetime64[ns](1), float64(7), int64(3), object(7)\n",
      "memory usage: 31.4+ MB\n"
     ]
    }
   ],
   "source": [
    "# Check datatypes of new columns\n",
    "raw_recipes_df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "80a458da",
   "metadata": {},
   "source": [
    "Wonderful! As a result of the data transformation, we have successfully created separate columns to represent each specific nutritional value. These columns now store the respective nutritional metrics as floating-point values, which facilitates further analysis and computations involving these nutritional attributes."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0e8b54c6",
   "metadata": {},
   "source": [
    "#### 4.1.2 Time Components <a class=\"anchor\" id=\"fouronetwo\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "54772be2",
   "metadata": {},
   "source": [
    "For both the recipe submission and recipe review, we will extract the month and year variables.\n",
    "\n",
    "Let us start with the recipe submission in the recipe dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "d26dae63",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['name', 'id', 'minutes', 'contributor_id', 'submitted', 'tags',\n",
       "       'n_steps', 'steps', 'description', 'ingredients', 'n_ingredients',\n",
       "       'calories', 'total_fat', 'sugar', 'sodium', 'protein', 'saturated_fat',\n",
       "       'carbohydrates'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# View date column name\n",
    "raw_recipes_df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "c1605212",
   "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>name</th>\n",
       "      <th>id</th>\n",
       "      <th>minutes</th>\n",
       "      <th>contributor_id</th>\n",
       "      <th>submitted</th>\n",
       "      <th>tags</th>\n",
       "      <th>n_steps</th>\n",
       "      <th>steps</th>\n",
       "      <th>description</th>\n",
       "      <th>ingredients</th>\n",
       "      <th>n_ingredients</th>\n",
       "      <th>calories</th>\n",
       "      <th>total_fat</th>\n",
       "      <th>sugar</th>\n",
       "      <th>sodium</th>\n",
       "      <th>protein</th>\n",
       "      <th>saturated_fat</th>\n",
       "      <th>carbohydrates</th>\n",
       "      <th>submitted_month</th>\n",
       "      <th>submitted_year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>arriba   baked winter squash mexican style</td>\n",
       "      <td>137739</td>\n",
       "      <td>55</td>\n",
       "      <td>47892</td>\n",
       "      <td>2005-09-16</td>\n",
       "      <td>['60-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>11</td>\n",
       "      <td>['make a choice and proceed with recipe', 'dep...</td>\n",
       "      <td>autumn is my favorite time of year to cook! th...</td>\n",
       "      <td>['winter squash', 'mexican seasoning', 'mixed ...</td>\n",
       "      <td>7</td>\n",
       "      <td>51.5</td>\n",
       "      <td>0.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>Sep</td>\n",
       "      <td>2005</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>a bit different  breakfast pizza</td>\n",
       "      <td>31490</td>\n",
       "      <td>30</td>\n",
       "      <td>26278</td>\n",
       "      <td>2002-06-17</td>\n",
       "      <td>['30-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>9</td>\n",
       "      <td>['preheat oven to 425 degrees f', 'press dough...</td>\n",
       "      <td>this recipe calls for the crust to be prebaked...</td>\n",
       "      <td>['prepared pizza crust', 'sausage patty', 'egg...</td>\n",
       "      <td>6</td>\n",
       "      <td>173.4</td>\n",
       "      <td>18.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>17.0</td>\n",
       "      <td>22.0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Jun</td>\n",
       "      <td>2002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>all in the kitchen  chili</td>\n",
       "      <td>112140</td>\n",
       "      <td>130</td>\n",
       "      <td>196586</td>\n",
       "      <td>2005-02-25</td>\n",
       "      <td>['time-to-make', 'course', 'preparation', 'mai...</td>\n",
       "      <td>6</td>\n",
       "      <td>['brown ground beef in large pot', 'add choppe...</td>\n",
       "      <td>this modified version of 'mom's' chili was a h...</td>\n",
       "      <td>['ground beef', 'yellow onions', 'diced tomato...</td>\n",
       "      <td>13</td>\n",
       "      <td>269.8</td>\n",
       "      <td>22.0</td>\n",
       "      <td>32.0</td>\n",
       "      <td>48.0</td>\n",
       "      <td>39.0</td>\n",
       "      <td>27.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>Feb</td>\n",
       "      <td>2005</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>alouette  potatoes</td>\n",
       "      <td>59389</td>\n",
       "      <td>45</td>\n",
       "      <td>68585</td>\n",
       "      <td>2003-04-14</td>\n",
       "      <td>['60-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>11</td>\n",
       "      <td>['place potatoes in a large pot of lightly sal...</td>\n",
       "      <td>this is a super easy, great tasting, make ahea...</td>\n",
       "      <td>['spreadable cheese with garlic and herbs', 'n...</td>\n",
       "      <td>11</td>\n",
       "      <td>368.1</td>\n",
       "      <td>17.0</td>\n",
       "      <td>10.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>20.0</td>\n",
       "      <td>Apr</td>\n",
       "      <td>2003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>amish  tomato ketchup  for canning</td>\n",
       "      <td>44061</td>\n",
       "      <td>190</td>\n",
       "      <td>41706</td>\n",
       "      <td>2002-10-25</td>\n",
       "      <td>['weeknight', 'time-to-make', 'course', 'main-...</td>\n",
       "      <td>5</td>\n",
       "      <td>['mix all ingredients&amp; boil for 2 1 / 2 hours ...</td>\n",
       "      <td>my dh's amish mother raised him on this recipe...</td>\n",
       "      <td>['tomato juice', 'apple cider vinegar', 'sugar...</td>\n",
       "      <td>8</td>\n",
       "      <td>352.9</td>\n",
       "      <td>1.0</td>\n",
       "      <td>337.0</td>\n",
       "      <td>23.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>28.0</td>\n",
       "      <td>Oct</td>\n",
       "      <td>2002</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         name      id  minutes contributor_id  \\\n",
       "0  arriba   baked winter squash mexican style  137739       55          47892   \n",
       "1            a bit different  breakfast pizza   31490       30          26278   \n",
       "2                   all in the kitchen  chili  112140      130         196586   \n",
       "3                          alouette  potatoes   59389       45          68585   \n",
       "4          amish  tomato ketchup  for canning   44061      190          41706   \n",
       "\n",
       "   submitted                                               tags  n_steps  \\\n",
       "0 2005-09-16  ['60-minutes-or-less', 'time-to-make', 'course...       11   \n",
       "1 2002-06-17  ['30-minutes-or-less', 'time-to-make', 'course...        9   \n",
       "2 2005-02-25  ['time-to-make', 'course', 'preparation', 'mai...        6   \n",
       "3 2003-04-14  ['60-minutes-or-less', 'time-to-make', 'course...       11   \n",
       "4 2002-10-25  ['weeknight', 'time-to-make', 'course', 'main-...        5   \n",
       "\n",
       "                                               steps  \\\n",
       "0  ['make a choice and proceed with recipe', 'dep...   \n",
       "1  ['preheat oven to 425 degrees f', 'press dough...   \n",
       "2  ['brown ground beef in large pot', 'add choppe...   \n",
       "3  ['place potatoes in a large pot of lightly sal...   \n",
       "4  ['mix all ingredients& boil for 2 1 / 2 hours ...   \n",
       "\n",
       "                                         description  \\\n",
       "0  autumn is my favorite time of year to cook! th...   \n",
       "1  this recipe calls for the crust to be prebaked...   \n",
       "2  this modified version of 'mom's' chili was a h...   \n",
       "3  this is a super easy, great tasting, make ahea...   \n",
       "4  my dh's amish mother raised him on this recipe...   \n",
       "\n",
       "                                         ingredients  n_ingredients  calories  \\\n",
       "0  ['winter squash', 'mexican seasoning', 'mixed ...              7      51.5   \n",
       "1  ['prepared pizza crust', 'sausage patty', 'egg...              6     173.4   \n",
       "2  ['ground beef', 'yellow onions', 'diced tomato...             13     269.8   \n",
       "3  ['spreadable cheese with garlic and herbs', 'n...             11     368.1   \n",
       "4  ['tomato juice', 'apple cider vinegar', 'sugar...              8     352.9   \n",
       "\n",
       "   total_fat  sugar  sodium  protein  saturated_fat  carbohydrates  \\\n",
       "0        0.0   13.0     0.0      2.0            0.0            4.0   \n",
       "1       18.0    0.0    17.0     22.0           35.0            1.0   \n",
       "2       22.0   32.0    48.0     39.0           27.0            5.0   \n",
       "3       17.0   10.0     2.0     14.0            8.0           20.0   \n",
       "4        1.0  337.0    23.0      3.0            0.0           28.0   \n",
       "\n",
       "  submitted_month  submitted_year  \n",
       "0             Sep            2005  \n",
       "1             Jun            2002  \n",
       "2             Feb            2005  \n",
       "3             Apr            2003  \n",
       "4             Oct            2002  "
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Define the month abbreviation mapping\n",
    "month_mapping = {\n",
    "    1: 'Jan',\n",
    "    2: 'Feb',\n",
    "    3: 'Mar',\n",
    "    4: 'Apr',\n",
    "    5: 'May',\n",
    "    6: 'Jun',\n",
    "    7: 'Jul',\n",
    "    8: 'Aug',\n",
    "    9: 'Sep',\n",
    "    10: 'Oct',\n",
    "    11: 'Nov',\n",
    "    12: 'Dec'\n",
    "}\n",
    "\n",
    "# Store month and year separately\n",
    "raw_recipes_df['submitted_month'] = raw_recipes_df['submitted'].dt.month.map(month_mapping)\n",
    "raw_recipes_df['submitted_year'] = raw_recipes_df['submitted'].dt.year\n",
    "\n",
    "# Sanity check\n",
    "raw_recipes_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "66000a0e",
   "metadata": {},
   "source": [
    "We will do the same for the date column in the user interactions dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "e437a19c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['user_id', 'recipe_id', 'date', 'rating', 'review'], dtype='object')"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# View date column name\n",
    "raw_users_df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "f3bea77a",
   "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>recipe_id</th>\n",
       "      <th>date</th>\n",
       "      <th>rating</th>\n",
       "      <th>review</th>\n",
       "      <th>review_month</th>\n",
       "      <th>review_year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>38094</td>\n",
       "      <td>40893</td>\n",
       "      <td>2003-02-17</td>\n",
       "      <td>4</td>\n",
       "      <td>Great with a salad. Cooked on top of stove for...</td>\n",
       "      <td>Feb</td>\n",
       "      <td>2003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1293707</td>\n",
       "      <td>40893</td>\n",
       "      <td>2011-12-21</td>\n",
       "      <td>5</td>\n",
       "      <td>So simple, so delicious! Great for chilly fall...</td>\n",
       "      <td>Dec</td>\n",
       "      <td>2011</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>8937</td>\n",
       "      <td>44394</td>\n",
       "      <td>2002-12-01</td>\n",
       "      <td>4</td>\n",
       "      <td>This worked very well and is EASY.  I used not...</td>\n",
       "      <td>Dec</td>\n",
       "      <td>2002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>126440</td>\n",
       "      <td>85009</td>\n",
       "      <td>2010-02-27</td>\n",
       "      <td>5</td>\n",
       "      <td>I made the Mexican topping and took it to bunk...</td>\n",
       "      <td>Feb</td>\n",
       "      <td>2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>57222</td>\n",
       "      <td>85009</td>\n",
       "      <td>2011-10-01</td>\n",
       "      <td>5</td>\n",
       "      <td>Made the cheddar bacon topping, adding a sprin...</td>\n",
       "      <td>Oct</td>\n",
       "      <td>2011</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   user_id recipe_id       date  rating  \\\n",
       "0    38094     40893 2003-02-17       4   \n",
       "1  1293707     40893 2011-12-21       5   \n",
       "2     8937     44394 2002-12-01       4   \n",
       "3   126440     85009 2010-02-27       5   \n",
       "4    57222     85009 2011-10-01       5   \n",
       "\n",
       "                                              review review_month  review_year  \n",
       "0  Great with a salad. Cooked on top of stove for...          Feb         2003  \n",
       "1  So simple, so delicious! Great for chilly fall...          Dec         2011  \n",
       "2  This worked very well and is EASY.  I used not...          Dec         2002  \n",
       "3  I made the Mexican topping and took it to bunk...          Feb         2010  \n",
       "4  Made the cheddar bacon topping, adding a sprin...          Oct         2011  "
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Store month and year separately\n",
    "raw_users_df['review_month'] = raw_users_df['date'].dt.month.map(month_mapping)\n",
    "raw_users_df['review_year'] = raw_users_df['date'].dt.year\n",
    "\n",
    "# Sanity check\n",
    "raw_users_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "35e6dfb3",
   "metadata": {},
   "source": [
    "#### 4.1.3 Dietary Preference <a class=\"anchor\" id=\"fouronethree\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d4f229c9",
   "metadata": {},
   "source": [
    "In this section, we will enrich our recipe dataset by creating five new features derived from the 'tags' column. These features will capture key dietary restrictions, namely dairy-free, gluten-free, low-carb, vegan, and vegetarian. By incorporating these dietary restriction features, we aim to enhance the personalization and relevance of our recipe recommendations."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "7828d4d1",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create 5 new features\n",
    "raw_recipes_df['dairy-free'] = raw_recipes_df['tags'].apply(lambda x: 1 if 'dairy-free' in x else 0)\n",
    "raw_recipes_df['gluten-free'] = raw_recipes_df['tags'].apply(lambda x: 1 if 'gluten-free' in x else 0)\n",
    "raw_recipes_df['low-carb'] = raw_recipes_df['tags'].apply(lambda x: 1 if 'low-carb' in x else 0)\n",
    "raw_recipes_df['vegan'] = raw_recipes_df['tags'].apply(lambda x: 1 if 'vegan' in x else 0)\n",
    "raw_recipes_df['vegetarian'] = raw_recipes_df['tags'].apply(lambda x: 1 if 'vegetarian' in x else 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "d8e4edd5",
   "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>name</th>\n",
       "      <th>id</th>\n",
       "      <th>minutes</th>\n",
       "      <th>contributor_id</th>\n",
       "      <th>submitted</th>\n",
       "      <th>tags</th>\n",
       "      <th>n_steps</th>\n",
       "      <th>steps</th>\n",
       "      <th>description</th>\n",
       "      <th>ingredients</th>\n",
       "      <th>...</th>\n",
       "      <th>protein</th>\n",
       "      <th>saturated_fat</th>\n",
       "      <th>carbohydrates</th>\n",
       "      <th>submitted_month</th>\n",
       "      <th>submitted_year</th>\n",
       "      <th>dairy-free</th>\n",
       "      <th>gluten-free</th>\n",
       "      <th>low-carb</th>\n",
       "      <th>vegan</th>\n",
       "      <th>vegetarian</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>172013</th>\n",
       "      <td>red curry hummus</td>\n",
       "      <td>395136</td>\n",
       "      <td>25</td>\n",
       "      <td>809163</td>\n",
       "      <td>2009-10-16</td>\n",
       "      <td>['weeknight', '30-minutes-or-less', 'time-to-m...</td>\n",
       "      <td>6</td>\n",
       "      <td>['prepare garbanzo beans by using fingers to r...</td>\n",
       "      <td>this one is fun!   i love adding unusual bold ...</td>\n",
       "      <td>['garbanzo beans', 'lemon juice', 'tahini past...</td>\n",
       "      <td>...</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>Oct</td>\n",
       "      <td>2009</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31719</th>\n",
       "      <td>butter baby</td>\n",
       "      <td>286448</td>\n",
       "      <td>5</td>\n",
       "      <td>327115</td>\n",
       "      <td>2008-02-13</td>\n",
       "      <td>['15-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>2</td>\n",
       "      <td>['combine ingredients in a beer mug', 'stir an...</td>\n",
       "      <td>this is delicious!  very sweet but fairly mild.</td>\n",
       "      <td>['coconut rum', 'amaretto', 'kahlua', 'butters...</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Feb</td>\n",
       "      <td>2008</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>187143</th>\n",
       "      <td>simple potato and rosemary focaccia</td>\n",
       "      <td>352681</td>\n",
       "      <td>110</td>\n",
       "      <td>323186</td>\n",
       "      <td>2009-01-28</td>\n",
       "      <td>['time-to-make', 'course', 'cuisine', 'prepara...</td>\n",
       "      <td>7</td>\n",
       "      <td>['in a large bowl , combine the mashed potatoe...</td>\n",
       "      <td>this was my first attempt at focaccia, several...</td>\n",
       "      <td>['mashed potatoes', 'white bread flour', 'yeas...</td>\n",
       "      <td>...</td>\n",
       "      <td>11.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>13.0</td>\n",
       "      <td>Jan</td>\n",
       "      <td>2009</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30855</th>\n",
       "      <td>bryan s chocolate lamb chili</td>\n",
       "      <td>96173</td>\n",
       "      <td>60</td>\n",
       "      <td>151679</td>\n",
       "      <td>2004-07-22</td>\n",
       "      <td>['60-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>7</td>\n",
       "      <td>['in a large pot , cook onions and ground lamb...</td>\n",
       "      <td>we love chili, we love chocolate, we love icel...</td>\n",
       "      <td>['onion', 'lean ground lamb', 'olive oil', 're...</td>\n",
       "      <td>...</td>\n",
       "      <td>46.0</td>\n",
       "      <td>43.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>Jul</td>\n",
       "      <td>2004</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16696</th>\n",
       "      <td>banana muffins with mascarpone cream frosting ...</td>\n",
       "      <td>130953</td>\n",
       "      <td>85</td>\n",
       "      <td>145352</td>\n",
       "      <td>2005-07-25</td>\n",
       "      <td>['time-to-make', 'course', 'main-ingredient', ...</td>\n",
       "      <td>14</td>\n",
       "      <td>['line 18 muffin cups with paper liners', 'pre...</td>\n",
       "      <td>these are from giada de laurentiis. so sinfull...</td>\n",
       "      <td>['all-purpose flour', 'baking soda', 'salt', '...</td>\n",
       "      <td>...</td>\n",
       "      <td>8.0</td>\n",
       "      <td>27.0</td>\n",
       "      <td>15.0</td>\n",
       "      <td>Jul</td>\n",
       "      <td>2005</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 25 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                     name      id  minutes  \\\n",
       "172013                                   red curry hummus  395136       25   \n",
       "31719                                         butter baby  286448        5   \n",
       "187143                simple potato and rosemary focaccia  352681      110   \n",
       "30855                        bryan s chocolate lamb chili   96173       60   \n",
       "16696   banana muffins with mascarpone cream frosting ...  130953       85   \n",
       "\n",
       "       contributor_id  submitted  \\\n",
       "172013         809163 2009-10-16   \n",
       "31719          327115 2008-02-13   \n",
       "187143         323186 2009-01-28   \n",
       "30855          151679 2004-07-22   \n",
       "16696          145352 2005-07-25   \n",
       "\n",
       "                                                     tags  n_steps  \\\n",
       "172013  ['weeknight', '30-minutes-or-less', 'time-to-m...        6   \n",
       "31719   ['15-minutes-or-less', 'time-to-make', 'course...        2   \n",
       "187143  ['time-to-make', 'course', 'cuisine', 'prepara...        7   \n",
       "30855   ['60-minutes-or-less', 'time-to-make', 'course...        7   \n",
       "16696   ['time-to-make', 'course', 'main-ingredient', ...       14   \n",
       "\n",
       "                                                    steps  \\\n",
       "172013  ['prepare garbanzo beans by using fingers to r...   \n",
       "31719   ['combine ingredients in a beer mug', 'stir an...   \n",
       "187143  ['in a large bowl , combine the mashed potatoe...   \n",
       "30855   ['in a large pot , cook onions and ground lamb...   \n",
       "16696   ['line 18 muffin cups with paper liners', 'pre...   \n",
       "\n",
       "                                              description  \\\n",
       "172013  this one is fun!   i love adding unusual bold ...   \n",
       "31719     this is delicious!  very sweet but fairly mild.   \n",
       "187143  this was my first attempt at focaccia, several...   \n",
       "30855   we love chili, we love chocolate, we love icel...   \n",
       "16696   these are from giada de laurentiis. so sinfull...   \n",
       "\n",
       "                                              ingredients  ...  protein  \\\n",
       "172013  ['garbanzo beans', 'lemon juice', 'tahini past...  ...      4.0   \n",
       "31719   ['coconut rum', 'amaretto', 'kahlua', 'butters...  ...      0.0   \n",
       "187143  ['mashed potatoes', 'white bread flour', 'yeas...  ...     11.0   \n",
       "30855   ['onion', 'lean ground lamb', 'olive oil', 're...  ...     46.0   \n",
       "16696   ['all-purpose flour', 'baking soda', 'salt', '...  ...      8.0   \n",
       "\n",
       "        saturated_fat  carbohydrates  submitted_month  submitted_year  \\\n",
       "172013            5.0            3.0              Oct            2009   \n",
       "31719             0.0            1.0              Feb            2008   \n",
       "187143            5.0           13.0              Jan            2009   \n",
       "30855            43.0           12.0              Jul            2004   \n",
       "16696            27.0           15.0              Jul            2005   \n",
       "\n",
       "        dairy-free  gluten-free  low-carb vegan  vegetarian  \n",
       "172013           0            0         0     0           0  \n",
       "31719            0            0         0     0           0  \n",
       "187143           0            0         0     0           1  \n",
       "30855            0            0         1     0           0  \n",
       "16696            0            0         0     0           0  \n",
       "\n",
       "[5 rows x 25 columns]"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Sanity check\n",
    "raw_recipes_df.sample(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "d38abb27",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 216818 entries, 0 to 231636\n",
      "Data columns (total 25 columns):\n",
      " #   Column           Non-Null Count   Dtype         \n",
      "---  ------           --------------   -----         \n",
      " 0   name             216818 non-null  object        \n",
      " 1   id               216818 non-null  object        \n",
      " 2   minutes          216818 non-null  int64         \n",
      " 3   contributor_id   216818 non-null  object        \n",
      " 4   submitted        216818 non-null  datetime64[ns]\n",
      " 5   tags             216818 non-null  object        \n",
      " 6   n_steps          216818 non-null  int64         \n",
      " 7   steps            216818 non-null  object        \n",
      " 8   description      216818 non-null  object        \n",
      " 9   ingredients      216818 non-null  object        \n",
      " 10  n_ingredients    216818 non-null  int64         \n",
      " 11  calories         216818 non-null  float64       \n",
      " 12  total_fat        216818 non-null  float64       \n",
      " 13  sugar            216818 non-null  float64       \n",
      " 14  sodium           216818 non-null  float64       \n",
      " 15  protein          216818 non-null  float64       \n",
      " 16  saturated_fat    216818 non-null  float64       \n",
      " 17  carbohydrates    216818 non-null  float64       \n",
      " 18  submitted_month  216818 non-null  object        \n",
      " 19  submitted_year   216818 non-null  int64         \n",
      " 20  dairy-free       216818 non-null  int64         \n",
      " 21  gluten-free      216818 non-null  int64         \n",
      " 22  low-carb         216818 non-null  int64         \n",
      " 23  vegan            216818 non-null  int64         \n",
      " 24  vegetarian       216818 non-null  int64         \n",
      "dtypes: datetime64[ns](1), float64(7), int64(9), object(8)\n",
      "memory usage: 43.0+ MB\n"
     ]
    }
   ],
   "source": [
    "# Another sanity check for newly-added columns\n",
    "raw_recipes_df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e94c5eaf",
   "metadata": {},
   "source": [
    "### 4.2 Aggregation <a class=\"anchor\" id=\"foutwo\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "419b9d99",
   "metadata": {},
   "source": [
    "To maintain the separation during the exploratory data analysis (EDA) phase, we will keep the recipe and user interaction datasets separate. However, when it comes to modeling, we can merge the two datasets and reduce the data to only the necessary information."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3e297d83",
   "metadata": {},
   "source": [
    "#### 4.2.1 Ratings <a class=\"anchor\" id=\"foutwoone\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d724d1ac",
   "metadata": {},
   "source": [
    "In this case, we will only require the 'rating' information from the user interactions data, which we will merge into the recipes dataset. This merging process will allow us to combine the relevant data from both datasets and enable us to build a comprehensive recipe recommendation model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "5c013c28",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['user_id', 'recipe_id', 'date', 'rating', 'review', 'review_month',\n",
       "       'review_year'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# View the column names\n",
    "raw_users_df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "bbf097c0",
   "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>recipe_id</th>\n",
       "      <th>rating</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>40893</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>40893</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>44394</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>85009</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>85009</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1132360</th>\n",
       "      <td>82303</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1132363</th>\n",
       "      <td>386618</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1132364</th>\n",
       "      <td>78003</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1132365</th>\n",
       "      <td>78003</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1132366</th>\n",
       "      <td>78003</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1071351 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        recipe_id  rating\n",
       "0           40893       4\n",
       "1           40893       5\n",
       "2           44394       4\n",
       "3           85009       5\n",
       "4           85009       5\n",
       "...           ...     ...\n",
       "1132360     82303       5\n",
       "1132363    386618       5\n",
       "1132364     78003       5\n",
       "1132365     78003       4\n",
       "1132366     78003       5\n",
       "\n",
       "[1071351 rows x 2 columns]"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Create a new df that contains the recipe_id and rating only\n",
    "user_rating = raw_users_df[['recipe_id', 'rating']].copy()\n",
    "user_rating"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "355b90c4",
   "metadata": {},
   "source": [
    "To obtain a holistic view of the average rating for each recipe, we will group the merged dataset by recipe and calculate the mean rating. This aggregation will provide us with a single representative rating value for each recipe, summarizing the overall sentiment and user satisfaction. By calculating the average rating, we can gain insights into the popularity and perceived quality of different recipes, which will be instrumental in our recipe recommendation system."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "2a256215",
   "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>recipe_id</th>\n",
       "      <th>mean</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>100</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>100000</td>\n",
       "      <td>4.708333</td>\n",
       "      <td>24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>100001</td>\n",
       "      <td>4.666667</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>100002</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  recipe_id      mean  count\n",
       "0       100  3.000000      2\n",
       "1     10000  3.000000      1\n",
       "2    100000  4.708333     24\n",
       "3    100001  4.666667      3\n",
       "4    100002  4.000000      1"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Group by recipe and take the count and the average rating\n",
    "user_rating = user_rating.groupby(['recipe_id'])['rating'].agg(['mean','count'])\n",
    "user_rating.reset_index(inplace=True)\n",
    "\n",
    "# View table\n",
    "user_rating.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "cfdf09b4",
   "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>recipe_id</th>\n",
       "      <th>average_rating</th>\n",
       "      <th>votes</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>100</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>100000</td>\n",
       "      <td>4.708333</td>\n",
       "      <td>24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>100001</td>\n",
       "      <td>4.666667</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>100002</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>226578</th>\n",
       "      <td>99995</td>\n",
       "      <td>4.545455</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>226579</th>\n",
       "      <td>99996</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>226580</th>\n",
       "      <td>99997</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>226581</th>\n",
       "      <td>99998</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>226582</th>\n",
       "      <td>99999</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>226583 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       recipe_id  average_rating  votes\n",
       "0            100        3.000000      2\n",
       "1          10000        3.000000      1\n",
       "2         100000        4.708333     24\n",
       "3         100001        4.666667      3\n",
       "4         100002        4.000000      1\n",
       "...          ...             ...    ...\n",
       "226578     99995        4.545455     11\n",
       "226579     99996        5.000000     11\n",
       "226580     99997        4.000000      1\n",
       "226581     99998        5.000000      2\n",
       "226582     99999        5.000000      1\n",
       "\n",
       "[226583 rows x 3 columns]"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Rename columns \n",
    "user_rating.rename(columns={'mean':'average_rating','count':'votes'},inplace=True)\n",
    "\n",
    "# View table\n",
    "user_rating"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0c0829df",
   "metadata": {},
   "source": [
    "#### 4.2.2 Merge Datasets <a class=\"anchor\" id=\"foutwotwo\"></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "9f68a085",
   "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>name</th>\n",
       "      <th>id</th>\n",
       "      <th>minutes</th>\n",
       "      <th>contributor_id</th>\n",
       "      <th>submitted</th>\n",
       "      <th>tags</th>\n",
       "      <th>n_steps</th>\n",
       "      <th>steps</th>\n",
       "      <th>description</th>\n",
       "      <th>ingredients</th>\n",
       "      <th>...</th>\n",
       "      <th>submitted_month</th>\n",
       "      <th>submitted_year</th>\n",
       "      <th>dairy-free</th>\n",
       "      <th>gluten-free</th>\n",
       "      <th>low-carb</th>\n",
       "      <th>vegan</th>\n",
       "      <th>vegetarian</th>\n",
       "      <th>recipe_id</th>\n",
       "      <th>average_rating</th>\n",
       "      <th>votes</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>arriba   baked winter squash mexican style</td>\n",
       "      <td>137739</td>\n",
       "      <td>55</td>\n",
       "      <td>47892</td>\n",
       "      <td>2005-09-16</td>\n",
       "      <td>['60-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>11</td>\n",
       "      <td>['make a choice and proceed with recipe', 'dep...</td>\n",
       "      <td>autumn is my favorite time of year to cook! th...</td>\n",
       "      <td>['winter squash', 'mexican seasoning', 'mixed ...</td>\n",
       "      <td>...</td>\n",
       "      <td>Sep</td>\n",
       "      <td>2005</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>137739</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>a bit different  breakfast pizza</td>\n",
       "      <td>31490</td>\n",
       "      <td>30</td>\n",
       "      <td>26278</td>\n",
       "      <td>2002-06-17</td>\n",
       "      <td>['30-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>9</td>\n",
       "      <td>['preheat oven to 425 degrees f', 'press dough...</td>\n",
       "      <td>this recipe calls for the crust to be prebaked...</td>\n",
       "      <td>['prepared pizza crust', 'sausage patty', 'egg...</td>\n",
       "      <td>...</td>\n",
       "      <td>Jun</td>\n",
       "      <td>2002</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>31490</td>\n",
       "      <td>4.666667</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>all in the kitchen  chili</td>\n",
       "      <td>112140</td>\n",
       "      <td>130</td>\n",
       "      <td>196586</td>\n",
       "      <td>2005-02-25</td>\n",
       "      <td>['time-to-make', 'course', 'preparation', 'mai...</td>\n",
       "      <td>6</td>\n",
       "      <td>['brown ground beef in large pot', 'add choppe...</td>\n",
       "      <td>this modified version of 'mom's' chili was a h...</td>\n",
       "      <td>['ground beef', 'yellow onions', 'diced tomato...</td>\n",
       "      <td>...</td>\n",
       "      <td>Feb</td>\n",
       "      <td>2005</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>112140</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>alouette  potatoes</td>\n",
       "      <td>59389</td>\n",
       "      <td>45</td>\n",
       "      <td>68585</td>\n",
       "      <td>2003-04-14</td>\n",
       "      <td>['60-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>11</td>\n",
       "      <td>['place potatoes in a large pot of lightly sal...</td>\n",
       "      <td>this is a super easy, great tasting, make ahea...</td>\n",
       "      <td>['spreadable cheese with garlic and herbs', 'n...</td>\n",
       "      <td>...</td>\n",
       "      <td>Apr</td>\n",
       "      <td>2003</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>59389</td>\n",
       "      <td>4.500000</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>amish  tomato ketchup  for canning</td>\n",
       "      <td>44061</td>\n",
       "      <td>190</td>\n",
       "      <td>41706</td>\n",
       "      <td>2002-10-25</td>\n",
       "      <td>['weeknight', 'time-to-make', 'course', 'main-...</td>\n",
       "      <td>5</td>\n",
       "      <td>['mix all ingredients&amp; boil for 2 1 / 2 hours ...</td>\n",
       "      <td>my dh's amish mother raised him on this recipe...</td>\n",
       "      <td>['tomato juice', 'apple cider vinegar', 'sugar...</td>\n",
       "      <td>...</td>\n",
       "      <td>Oct</td>\n",
       "      <td>2002</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>44061</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>212250</th>\n",
       "      <td>zydeco soup</td>\n",
       "      <td>486161</td>\n",
       "      <td>60</td>\n",
       "      <td>227978</td>\n",
       "      <td>2012-08-29</td>\n",
       "      <td>['ham', '60-minutes-or-less', 'time-to-make', ...</td>\n",
       "      <td>7</td>\n",
       "      <td>['heat oil in a 4-quart dutch oven', 'add cele...</td>\n",
       "      <td>this is a delicious soup that i originally fou...</td>\n",
       "      <td>['celery', 'onion', 'green sweet pepper', 'gar...</td>\n",
       "      <td>...</td>\n",
       "      <td>Aug</td>\n",
       "      <td>2012</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>486161</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>212251</th>\n",
       "      <td>zydeco spice mix</td>\n",
       "      <td>493372</td>\n",
       "      <td>5</td>\n",
       "      <td>1500678</td>\n",
       "      <td>2013-01-09</td>\n",
       "      <td>['15-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>1</td>\n",
       "      <td>['mix all ingredients together thoroughly']</td>\n",
       "      <td>this spice mix will make your taste buds dance!</td>\n",
       "      <td>['paprika', 'salt', 'garlic powder', 'onion po...</td>\n",
       "      <td>...</td>\n",
       "      <td>Jan</td>\n",
       "      <td>2013</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>493372</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>212252</th>\n",
       "      <td>zydeco ya ya deviled eggs</td>\n",
       "      <td>308080</td>\n",
       "      <td>40</td>\n",
       "      <td>37779</td>\n",
       "      <td>2008-06-07</td>\n",
       "      <td>['60-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>7</td>\n",
       "      <td>['in a bowl , combine the mashed yolks and may...</td>\n",
       "      <td>deviled eggs, cajun-style</td>\n",
       "      <td>['hard-cooked eggs', 'mayonnaise', 'dijon must...</td>\n",
       "      <td>...</td>\n",
       "      <td>Jun</td>\n",
       "      <td>2008</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>308080</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>212253</th>\n",
       "      <td>cookies by design   cookies on a stick</td>\n",
       "      <td>298512</td>\n",
       "      <td>29</td>\n",
       "      <td>506822</td>\n",
       "      <td>2008-04-15</td>\n",
       "      <td>['30-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>9</td>\n",
       "      <td>['place melted butter in a large mixing bowl a...</td>\n",
       "      <td>i've heard of the 'cookies by design' company,...</td>\n",
       "      <td>['butter', 'eagle brand condensed milk', 'ligh...</td>\n",
       "      <td>...</td>\n",
       "      <td>Apr</td>\n",
       "      <td>2008</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>298512</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>212254</th>\n",
       "      <td>cookies by design   sugar shortbread cookies</td>\n",
       "      <td>298509</td>\n",
       "      <td>20</td>\n",
       "      <td>506822</td>\n",
       "      <td>2008-04-15</td>\n",
       "      <td>['30-minutes-or-less', 'time-to-make', 'course...</td>\n",
       "      <td>5</td>\n",
       "      <td>['whip sugar and shortening in a large bowl , ...</td>\n",
       "      <td>i've heard of the 'cookies by design' company,...</td>\n",
       "      <td>['granulated sugar', 'shortening', 'eggs', 'fl...</td>\n",
       "      <td>...</td>\n",
       "      <td>Apr</td>\n",
       "      <td>2008</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>298509</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>212255 rows × 28 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                name      id  minutes  \\\n",
       "0         arriba   baked winter squash mexican style  137739       55   \n",
       "1                   a bit different  breakfast pizza   31490       30   \n",
       "2                          all in the kitchen  chili  112140      130   \n",
       "3                                 alouette  potatoes   59389       45   \n",
       "4                 amish  tomato ketchup  for canning   44061      190   \n",
       "...                                              ...     ...      ...   \n",
       "212250                                   zydeco soup  486161       60   \n",
       "212251                              zydeco spice mix  493372        5   \n",
       "212252                     zydeco ya ya deviled eggs  308080       40   \n",
       "212253        cookies by design   cookies on a stick  298512       29   \n",
       "212254  cookies by design   sugar shortbread cookies  298509       20   \n",
       "\n",
       "       contributor_id  submitted  \\\n",
       "0               47892 2005-09-16   \n",
       "1               26278 2002-06-17   \n",
       "2              196586 2005-02-25   \n",
       "3               68585 2003-04-14   \n",
       "4               41706 2002-10-25   \n",
       "...               ...        ...   \n",
       "212250         227978 2012-08-29   \n",
       "212251        1500678 2013-01-09   \n",
       "212252          37779 2008-06-07   \n",
       "212253         506822 2008-04-15   \n",
       "212254         506822 2008-04-15   \n",
       "\n",
       "                                                     tags  n_steps  \\\n",
       "0       ['60-minutes-or-less', 'time-to-make', 'course...       11   \n",
       "1       ['30-minutes-or-less', 'time-to-make', 'course...        9   \n",
       "2       ['time-to-make', 'course', 'preparation', 'mai...        6   \n",
       "3       ['60-minutes-or-less', 'time-to-make', 'course...       11   \n",
       "4       ['weeknight', 'time-to-make', 'course', 'main-...        5   \n",
       "...                                                   ...      ...   \n",
       "212250  ['ham', '60-minutes-or-less', 'time-to-make', ...        7   \n",
       "212251  ['15-minutes-or-less', 'time-to-make', 'course...        1   \n",
       "212252  ['60-minutes-or-less', 'time-to-make', 'course...        7   \n",
       "212253  ['30-minutes-or-less', 'time-to-make', 'course...        9   \n",
       "212254  ['30-minutes-or-less', 'time-to-make', 'course...        5   \n",
       "\n",
       "                                                    steps  \\\n",
       "0       ['make a choice and proceed with recipe', 'dep...   \n",
       "1       ['preheat oven to 425 degrees f', 'press dough...   \n",
       "2       ['brown ground beef in large pot', 'add choppe...   \n",
       "3       ['place potatoes in a large pot of lightly sal...   \n",
       "4       ['mix all ingredients& boil for 2 1 / 2 hours ...   \n",
       "...                                                   ...   \n",
       "212250  ['heat oil in a 4-quart dutch oven', 'add cele...   \n",
       "212251        ['mix all ingredients together thoroughly']   \n",
       "212252  ['in a bowl , combine the mashed yolks and may...   \n",
       "212253  ['place melted butter in a large mixing bowl a...   \n",
       "212254  ['whip sugar and shortening in a large bowl , ...   \n",
       "\n",
       "                                              description  \\\n",
       "0       autumn is my favorite time of year to cook! th...   \n",
       "1       this recipe calls for the crust to be prebaked...   \n",
       "2       this modified version of 'mom's' chili was a h...   \n",
       "3       this is a super easy, great tasting, make ahea...   \n",
       "4       my dh's amish mother raised him on this recipe...   \n",
       "...                                                   ...   \n",
       "212250  this is a delicious soup that i originally fou...   \n",
       "212251    this spice mix will make your taste buds dance!   \n",
       "212252                          deviled eggs, cajun-style   \n",
       "212253  i've heard of the 'cookies by design' company,...   \n",
       "212254  i've heard of the 'cookies by design' company,...   \n",
       "\n",
       "                                              ingredients  ...  \\\n",
       "0       ['winter squash', 'mexican seasoning', 'mixed ...  ...   \n",
       "1       ['prepared pizza crust', 'sausage patty', 'egg...  ...   \n",
       "2       ['ground beef', 'yellow onions', 'diced tomato...  ...   \n",
       "3       ['spreadable cheese with garlic and herbs', 'n...  ...   \n",
       "4       ['tomato juice', 'apple cider vinegar', 'sugar...  ...   \n",
       "...                                                   ...  ...   \n",
       "212250  ['celery', 'onion', 'green sweet pepper', 'gar...  ...   \n",
       "212251  ['paprika', 'salt', 'garlic powder', 'onion po...  ...   \n",
       "212252  ['hard-cooked eggs', 'mayonnaise', 'dijon must...  ...   \n",
       "212253  ['butter', 'eagle brand condensed milk', 'ligh...  ...   \n",
       "212254  ['granulated sugar', 'shortening', 'eggs', 'fl...  ...   \n",
       "\n",
       "        submitted_month  submitted_year  dairy-free  gluten-free  low-carb  \\\n",
       "0                   Sep            2005           0            0         0   \n",
       "1                   Jun            2002           0            0         0   \n",
       "2                   Feb            2005           0            0         0   \n",
       "3                   Apr            2003           0            0         0   \n",
       "4                   Oct            2002           0            0         0   \n",
       "...                 ...             ...         ...          ...       ...   \n",
       "212250              Aug            2012           0            0         0   \n",
       "212251              Jan            2013           0            0         0   \n",
       "212252              Jun            2008           0            0         0   \n",
       "212253              Apr            2008           0            0         0   \n",
       "212254              Apr            2008           0            0         0   \n",
       "\n",
       "        vegan  vegetarian  recipe_id average_rating  votes  \n",
       "0           0           1     137739       5.000000      3  \n",
       "1           0           0      31490       4.666667      3  \n",
       "2           0           0     112140       4.000000      1  \n",
       "3           0           0      59389       4.500000      2  \n",
       "4           0           1      44061       5.000000      1  \n",
       "...       ...         ...        ...            ...    ...  \n",
       "212250      0           0     486161       5.000000      6  \n",
       "212251      1           1     493372       5.000000      1  \n",
       "212252      0           0     308080       5.000000      5  \n",
       "212253      0           0     298512       1.000000      1  \n",
       "212254      0           0     298509       3.000000      2  \n",
       "\n",
       "[212255 rows x 28 columns]"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Merge ratings dataframe with recipe Dataframe\n",
    "food_df = pd.merge(raw_recipes_df, user_rating, left_on='id',right_on='recipe_id')\n",
    "\n",
    "# View table\n",
    "food_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "b40e7fa3",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 212255 entries, 0 to 212254\n",
      "Data columns (total 28 columns):\n",
      " #   Column           Non-Null Count   Dtype         \n",
      "---  ------           --------------   -----         \n",
      " 0   name             212255 non-null  object        \n",
      " 1   id               212255 non-null  object        \n",
      " 2   minutes          212255 non-null  int64         \n",
      " 3   contributor_id   212255 non-null  object        \n",
      " 4   submitted        212255 non-null  datetime64[ns]\n",
      " 5   tags             212255 non-null  object        \n",
      " 6   n_steps          212255 non-null  int64         \n",
      " 7   steps            212255 non-null  object        \n",
      " 8   description      212255 non-null  object        \n",
      " 9   ingredients      212255 non-null  object        \n",
      " 10  n_ingredients    212255 non-null  int64         \n",
      " 11  calories         212255 non-null  float64       \n",
      " 12  total_fat        212255 non-null  float64       \n",
      " 13  sugar            212255 non-null  float64       \n",
      " 14  sodium           212255 non-null  float64       \n",
      " 15  protein          212255 non-null  float64       \n",
      " 16  saturated_fat    212255 non-null  float64       \n",
      " 17  carbohydrates    212255 non-null  float64       \n",
      " 18  submitted_month  212255 non-null  object        \n",
      " 19  submitted_year   212255 non-null  int64         \n",
      " 20  dairy-free       212255 non-null  int64         \n",
      " 21  gluten-free      212255 non-null  int64         \n",
      " 22  low-carb         212255 non-null  int64         \n",
      " 23  vegan            212255 non-null  int64         \n",
      " 24  vegetarian       212255 non-null  int64         \n",
      " 25  recipe_id        212255 non-null  object        \n",
      " 26  average_rating   212255 non-null  float64       \n",
      " 27  votes            212255 non-null  int64         \n",
      "dtypes: datetime64[ns](1), float64(8), int64(10), object(9)\n",
      "memory usage: 47.0+ MB\n"
     ]
    }
   ],
   "source": [
    "# Final sanity check\n",
    "food_df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c00f9db7",
   "metadata": {},
   "source": [
    "#### Saving the data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "02a2c9c4",
   "metadata": {},
   "source": [
    "After diligently cleaning the datasets, it is important to save them to files for future use. By preserving these clean datasets, we can readily leverage them during the exploratory data analysis (EDA) phase. Saving the cleaned datasets ensures that the processed data is readily available, allowing us to delve into a comprehensive analysis of the dataset's characteristics and relationships."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "d391e92f",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Save the clean datasets\n",
    "\n",
    "# Individual ones for EDA\n",
    "raw_recipes_df.to_pickle(\"clean_recipes.pkl\")\n",
    "raw_users_df.to_pickle(\"clean_interactions.pkl\")\n",
    "# Merged for modelling\n",
    "food_df.to_pickle(\"food.pkl\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "023cbef5",
   "metadata": {},
   "source": [
    "## 5 Conclusion <a class=\"anchor\" id=\"five\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "97c40f0a",
   "metadata": {},
   "source": [
    "During the data preprocessing step of the recipe recommendation system, the following changes and actions were performed:\n",
    "\n",
    "- Converted the 'id' columns from integer to string data type for consistency and manipulation ease.\n",
    "- Dropped rows with missing values in the 'name' column and rows with missing values in the 'description' column.\n",
    "- Verified that there were no duplicate rows in the dataset.\n",
    "- Split the values in the 'nutrition' column, creating new columns for each individual nutritional value (calories, total fat, sugar, sodium, protein, saturated fat, carbohydrates).\n",
    "- Extracted the month and year from the 'submitted' and 'date' columns and created 4 new features.\n",
    "- Extracted dietary preferences from the 'tags' column and created 5 new features (dairy-free, gluten-free, low-carb, vegan, vegetarian).\n",
    "- Merged the recipes and user interactions datasets.\n",
    "\n",
    "With the datasets now clean and appropriately processed, we are well-prepared for the next phase of our project, which involves EDA and modelling. "
   ]
  }
 ],
 "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": "3.8.17"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}