model/06_Random_Forest_oversampling_unigram.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[nltk_data] Downloading package punkt to\n",
      "[nltk_data]     C:\\Users\\Boss\\AppData\\Roaming\\nltk_data...\n",
      "[nltk_data]   Package punkt is already up-to-date!\n",
      "[nltk_data] Downloading package stopwords to\n",
      "[nltk_data]     C:\\Users\\Boss\\AppData\\Roaming\\nltk_data...\n",
      "[nltk_data]   Package stopwords is already up-to-date!\n"
     ]
    }
   ],
   "source": [
    "# Import dependencies\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "\n",
    "# Plot confusion matrices\n",
    "import matplotlib as mpl \n",
    "import matplotlib.cm as cm \n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# NLP libraries\n",
    "import re\n",
    "import string\n",
    "import unicodedata\n",
    "import nltk\n",
    "nltk.download('punkt')\n",
    "from nltk.tokenize import word_tokenize\n",
    "from nltk.corpus import stopwords\n",
    "nltk.download('stopwords')\n",
    "stop_words_nltk = set(stopwords.words('english'))\n",
    "from nltk.stem.porter import PorterStemmer\n",
    "stemmer = PorterStemmer()\n",
    "\n",
    "# Machine learning\n",
    "import sklearn\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "from sklearn.utils import resample\n",
    "\n",
    "# Evaluate the classifiers\n",
    "from sklearn import metrics\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.metrics import confusion_matrix\n",
    "\n",
    "# Import time function from time module to track the training duration\n",
    "from time import time"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Connect to database"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# SQL Alchemy\n",
    "from sqlalchemy import create_engine"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create engine and connection\n",
    "engine = create_engine(\"sqlite:///../db.sqlite\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['cuisine_ingredients']"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check for tables\n",
    "engine.table_names()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Read in the data table\n",
    "df = pd.read_sql_query('SELECT * FROM cuisine_ingredients', con=engine)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>cuisine</th>\n",
       "      <th>recipe</th>\n",
       "      <th>full_ingredients</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3727</th>\n",
       "      <td>3728</td>\n",
       "      <td>Italian</td>\n",
       "      <td>Spicy fregola with scallops and crispy kale</td>\n",
       "      <td>['olive oil, for shallow and deep frying', '1 ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1923</th>\n",
       "      <td>1924</td>\n",
       "      <td>Chinese</td>\n",
       "      <td>Chunky bacon and cucumber salad</td>\n",
       "      <td>['1 tbsp groundnut oil', '6 long dried chillie...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3552</th>\n",
       "      <td>3553</td>\n",
       "      <td>Italian</td>\n",
       "      <td>Polenta pork</td>\n",
       "      <td>['2 tbsp sunflower oil ', '12 rashers thick ri...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1146</th>\n",
       "      <td>1147</td>\n",
       "      <td>British</td>\n",
       "      <td>Oat fig stuffing</td>\n",
       "      <td>['140g/5oz butter', '100g/3½oz jumbo oats', '1...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3195</th>\n",
       "      <td>3196</td>\n",
       "      <td>Italian</td>\n",
       "      <td>Bolognese with tagliatelle</td>\n",
       "      <td>['250g/9oz 00 flour', '3-4 medium free-range e...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1643</th>\n",
       "      <td>1644</td>\n",
       "      <td>British</td>\n",
       "      <td>Textured potato salad</td>\n",
       "      <td>['675g/1½lb floury potatoes, peeled and cut in...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3054</th>\n",
       "      <td>3055</td>\n",
       "      <td>Indian</td>\n",
       "      <td>Tandoori lamb wrap</td>\n",
       "      <td>['150ml/5fl oz Greek-style yoghurt ', '½ small...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>842</th>\n",
       "      <td>843</td>\n",
       "      <td>British</td>\n",
       "      <td>Gammon and pease pudding</td>\n",
       "      <td>['300g/10oz dried yellow split peas', '50g/2oz...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>505</th>\n",
       "      <td>506</td>\n",
       "      <td>British</td>\n",
       "      <td>Braised shin of beef with parsnip purée</td>\n",
       "      <td>['4kg/9lb beef shin, on the bone', 'sea salt a...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2224</th>\n",
       "      <td>2225</td>\n",
       "      <td>French</td>\n",
       "      <td>Mary Berry's celeriac remoulade</td>\n",
       "      <td>['650g/1lb 7oz celeriac, peeled and sliced int...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        id  cuisine                                       recipe  \\\n",
       "3727  3728  Italian  Spicy fregola with scallops and crispy kale   \n",
       "1923  1924  Chinese              Chunky bacon and cucumber salad   \n",
       "3552  3553  Italian                                Polenta pork    \n",
       "1146  1147  British                             Oat fig stuffing   \n",
       "3195  3196  Italian                   Bolognese with tagliatelle   \n",
       "1643  1644  British                        Textured potato salad   \n",
       "3054  3055   Indian                           Tandoori lamb wrap   \n",
       "842    843  British                     Gammon and pease pudding   \n",
       "505    506  British      Braised shin of beef with parsnip purée   \n",
       "2224  2225   French              Mary Berry's celeriac remoulade   \n",
       "\n",
       "                                       full_ingredients  \n",
       "3727  ['olive oil, for shallow and deep frying', '1 ...  \n",
       "1923  ['1 tbsp groundnut oil', '6 long dried chillie...  \n",
       "3552  ['2 tbsp sunflower oil ', '12 rashers thick ri...  \n",
       "1146  ['140g/5oz butter', '100g/3½oz jumbo oats', '1...  \n",
       "3195  ['250g/9oz 00 flour', '3-4 medium free-range e...  \n",
       "1643  ['675g/1½lb floury potatoes, peeled and cut in...  \n",
       "3054  ['150ml/5fl oz Greek-style yoghurt ', '½ small...  \n",
       "842   ['300g/10oz dried yellow split peas', '50g/2oz...  \n",
       "505   ['4kg/9lb beef shin, on the bone', 'sea salt a...  \n",
       "2224  ['650g/1lb 7oz celeriac, peeled and sliced int...  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Overview of the data set\n",
    "df.sample(10, random_state=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['African',\n",
       " 'American',\n",
       " 'British',\n",
       " 'Caribbean',\n",
       " 'Chinese',\n",
       " 'East European',\n",
       " 'French',\n",
       " 'Greek',\n",
       " 'Indian',\n",
       " 'Irish',\n",
       " 'Italian',\n",
       " 'Japanese',\n",
       " 'Korean',\n",
       " 'Mexican',\n",
       " 'Nordic',\n",
       " 'North African',\n",
       " 'Pakistani',\n",
       " 'Portuguese',\n",
       " 'South American',\n",
       " 'Spanish',\n",
       " 'Thai and South-East Asian',\n",
       " 'Turkish and Middle Eastern']"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get a list of unique cuisines in the dataset\n",
    "cuisine_list = df.cuisine.unique().tolist()\n",
    "cuisine_list"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Define text pre-processing function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create a function to remove accented characters\n",
    "def remove_accented_chars(matchobj):\n",
    "    text = matchobj.group()\n",
    "    new_text = unicodedata.normalize('NFKD', text).encode('ascii', 'ignore').decode('utf-8', 'ignore')\n",
    "    return new_text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "words_to_remove = [\n",
    "    \"tbsp\", \"roughly\", \"chopped\", \"tsp\", \"finely\", \"oz\", \"plus\", \"optional\",\n",
    "    \"extra\", \"fresh\", \"freshly\", \"ground\", \"thinly\", \"sliced\", \"clove\", \"pint\",\n",
    "    \"cut\", \"kg\", \"lb\", \"cm\", \"ml\", \"mm\", \"small\", \"large\", \"medium\", \"diced\", \"slice\",\n",
    "    \"pinch\", \"peeled\", \"grated\", \"removed\", \"handful\", \"piece\", \"crushed\", \"red\", \"dried\",\n",
    "    \"drained\", \"rinsed\", \"halved\", \"trimmed\", \"deseeded\", \"x\", \"beaten\", \"available\", \"supermarket\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create a function to clean ingredient text\n",
    "def clean(doc):\n",
    "    doc = doc.str.lower()\n",
    "    doc = doc.str.replace(r'\\w*[\\d¼½¾⅓⅔⅛⅜⅝]\\w*', '')\n",
    "    doc = doc.str.translate(str.maketrans('', '', string.punctuation))\n",
    "    doc = doc.str.replace(r'[£×–‘’“”⁄]', '')\n",
    "    doc = doc.apply(lambda x: re.sub(r'[âãäçèéêîïñóôûüōưấớ]', remove_accented_chars, x))\n",
    "    doc = doc.apply(lambda x: word_tokenize(x))\n",
    "    doc = doc.apply(lambda x: [word for word in x if not word in stop_words_nltk])\n",
    "    doc = doc.apply(lambda x: [word for word in x if not word in words_to_remove])\n",
    "    doc = doc.apply(lambda x: [stemmer.stem(word) for word in x])\n",
    "    processed_doc = doc.apply(lambda x: ' '.join([word for word in x]))\n",
    "\n",
    "    return processed_doc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Add a new column to the dataframe with the cleaned text\n",
    "df[\"ingredients_processed\"] = clean(df.full_ingredients)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>cuisine</th>\n",
       "      <th>recipe</th>\n",
       "      <th>full_ingredients</th>\n",
       "      <th>ingredients_processed</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3727</th>\n",
       "      <td>3728</td>\n",
       "      <td>Italian</td>\n",
       "      <td>Spicy fregola with scallops and crispy kale</td>\n",
       "      <td>['olive oil, for shallow and deep frying', '1 ...</td>\n",
       "      <td>oliv oil shallow deep fri shallot garlic chill...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1923</th>\n",
       "      <td>1924</td>\n",
       "      <td>Chinese</td>\n",
       "      <td>Chunky bacon and cucumber salad</td>\n",
       "      <td>['1 tbsp groundnut oil', '6 long dried chillie...</td>\n",
       "      <td>groundnut oil long chilli sichuan peppercorn w...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3552</th>\n",
       "      <td>3553</td>\n",
       "      <td>Italian</td>\n",
       "      <td>Polenta pork</td>\n",
       "      <td>['2 tbsp sunflower oil ', '12 rashers thick ri...</td>\n",
       "      <td>sunflow oil rasher thick rindless smoke streak...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1146</th>\n",
       "      <td>1147</td>\n",
       "      <td>British</td>\n",
       "      <td>Oat fig stuffing</td>\n",
       "      <td>['140g/5oz butter', '100g/3½oz jumbo oats', '1...</td>\n",
       "      <td>butter jumbo oat roll oat onion pistachio fig ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3195</th>\n",
       "      <td>3196</td>\n",
       "      <td>Italian</td>\n",
       "      <td>Bolognese with tagliatelle</td>\n",
       "      <td>['250g/9oz 00 flour', '3-4 medium free-range e...</td>\n",
       "      <td>flour freerang egg prefer bright yellow yolk f...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1643</th>\n",
       "      <td>1644</td>\n",
       "      <td>British</td>\n",
       "      <td>Textured potato salad</td>\n",
       "      <td>['675g/1½lb floury potatoes, peeled and cut in...</td>\n",
       "      <td>flouri potato cube rasher smoke streaki bacon ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3054</th>\n",
       "      <td>3055</td>\n",
       "      <td>Indian</td>\n",
       "      <td>Tandoori lamb wrap</td>\n",
       "      <td>['150ml/5fl oz Greek-style yoghurt ', '½ small...</td>\n",
       "      <td>greekstyl yoghurt onion garlic ginger garam ma...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>842</th>\n",
       "      <td>843</td>\n",
       "      <td>British</td>\n",
       "      <td>Gammon and pease pudding</td>\n",
       "      <td>['300g/10oz dried yellow split peas', '50g/2oz...</td>\n",
       "      <td>yellow split pea butter onion thyme bay leaf s...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>505</th>\n",
       "      <td>506</td>\n",
       "      <td>British</td>\n",
       "      <td>Braised shin of beef with parsnip purée</td>\n",
       "      <td>['4kg/9lb beef shin, on the bone', 'sea salt a...</td>\n",
       "      <td>beef shin bone sea salt black pepper oliv oil ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2224</th>\n",
       "      <td>2225</td>\n",
       "      <td>French</td>\n",
       "      <td>Mary Berry's celeriac remoulade</td>\n",
       "      <td>['650g/1lb 7oz celeriac, peeled and sliced int...</td>\n",
       "      <td>celeriac thin matchstick see tip lemon juic li...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        id  cuisine                                       recipe  \\\n",
       "3727  3728  Italian  Spicy fregola with scallops and crispy kale   \n",
       "1923  1924  Chinese              Chunky bacon and cucumber salad   \n",
       "3552  3553  Italian                                Polenta pork    \n",
       "1146  1147  British                             Oat fig stuffing   \n",
       "3195  3196  Italian                   Bolognese with tagliatelle   \n",
       "1643  1644  British                        Textured potato salad   \n",
       "3054  3055   Indian                           Tandoori lamb wrap   \n",
       "842    843  British                     Gammon and pease pudding   \n",
       "505    506  British      Braised shin of beef with parsnip purée   \n",
       "2224  2225   French              Mary Berry's celeriac remoulade   \n",
       "\n",
       "                                       full_ingredients  \\\n",
       "3727  ['olive oil, for shallow and deep frying', '1 ...   \n",
       "1923  ['1 tbsp groundnut oil', '6 long dried chillie...   \n",
       "3552  ['2 tbsp sunflower oil ', '12 rashers thick ri...   \n",
       "1146  ['140g/5oz butter', '100g/3½oz jumbo oats', '1...   \n",
       "3195  ['250g/9oz 00 flour', '3-4 medium free-range e...   \n",
       "1643  ['675g/1½lb floury potatoes, peeled and cut in...   \n",
       "3054  ['150ml/5fl oz Greek-style yoghurt ', '½ small...   \n",
       "842   ['300g/10oz dried yellow split peas', '50g/2oz...   \n",
       "505   ['4kg/9lb beef shin, on the bone', 'sea salt a...   \n",
       "2224  ['650g/1lb 7oz celeriac, peeled and sliced int...   \n",
       "\n",
       "                                  ingredients_processed  \n",
       "3727  oliv oil shallow deep fri shallot garlic chill...  \n",
       "1923  groundnut oil long chilli sichuan peppercorn w...  \n",
       "3552  sunflow oil rasher thick rindless smoke streak...  \n",
       "1146  butter jumbo oat roll oat onion pistachio fig ...  \n",
       "3195  flour freerang egg prefer bright yellow yolk f...  \n",
       "1643  flouri potato cube rasher smoke streaki bacon ...  \n",
       "3054  greekstyl yoghurt onion garlic ginger garam ma...  \n",
       "842   yellow split pea butter onion thyme bay leaf s...  \n",
       "505   beef shin bone sea salt black pepper oliv oil ...  \n",
       "2224  celeriac thin matchstick see tip lemon juic li...  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Overview of the dataset\n",
    "df.sample(10, random_state=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Modeling"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Split the data into training and test sets (75% train and 25% test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(4724,) (4724,)\n"
     ]
    }
   ],
   "source": [
    "# The column contains textual data to extract features from.\n",
    "X = df.ingredients_processed\n",
    "\n",
    "# The column we're learning to predict.\n",
    "y = df.cuisine \n",
    "print(X.shape, y.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3543,) (3543,)\n",
      "(1181,) (1181,)\n"
     ]
    }
   ],
   "source": [
    "# Split X and y into training and testing sets. By default, it splits 75% training and 25% test\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=1)\n",
    "print(X_train.shape, y_train.shape)\n",
    "print(X_test.shape, y_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Resampling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3543, 2)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Concatenate our training data back together\n",
    "X_y_train = pd.concat([X_train, y_train], axis=1)\n",
    "X_y_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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>ingredients_processed</th>\n",
       "      <th>cuisine</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1564</th>\n",
       "      <td>stew steak chunk beef kidney chunk season flou...</td>\n",
       "      <td>British</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3569</th>\n",
       "      <td>whole squid tube wing tentacl reserv oliv oil ...</td>\n",
       "      <td>Italian</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>177</th>\n",
       "      <td>freerang egg milk plain flour frozen spinach d...</td>\n",
       "      <td>American</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018</th>\n",
       "      <td>beef short rib bone neutral oil groundnut fri ...</td>\n",
       "      <td>Chinese</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2772</th>\n",
       "      <td>freerang egg light oliv oil cumin seed black m...</td>\n",
       "      <td>Indian</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  ingredients_processed   cuisine\n",
       "1564  stew steak chunk beef kidney chunk season flou...   British\n",
       "3569  whole squid tube wing tentacl reserv oliv oil ...   Italian\n",
       "177   freerang egg milk plain flour frozen spinach d...  American\n",
       "2018  beef short rib bone neutral oil groundnut fri ...   Chinese\n",
       "2772  freerang egg light oliv oil cumin seed black m...    Indian"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Overview of the training set\n",
    "X_y_train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ingredients_processed</th>\n",
       "      <th>cuisine</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3569</th>\n",
       "      <td>whole squid tube wing tentacl reserv oliv oil ...</td>\n",
       "      <td>Italian</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>177</th>\n",
       "      <td>freerang egg milk plain flour frozen spinach d...</td>\n",
       "      <td>American</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018</th>\n",
       "      <td>beef short rib bone neutral oil groundnut fri ...</td>\n",
       "      <td>Chinese</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2772</th>\n",
       "      <td>freerang egg light oliv oil cumin seed black m...</td>\n",
       "      <td>Indian</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3732</th>\n",
       "      <td>wholem penn veget oil onion garlic clove sweet...</td>\n",
       "      <td>Italian</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  ingredients_processed   cuisine\n",
       "3569  whole squid tube wing tentacl reserv oliv oil ...   Italian\n",
       "177   freerang egg milk plain flour frozen spinach d...  American\n",
       "2018  beef short rib bone neutral oil groundnut fri ...   Chinese\n",
       "2772  freerang egg light oliv oil cumin seed black m...    Indian\n",
       "3732  wholem penn veget oil onion garlic clove sweet...   Italian"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Separate minority and majority classes\n",
    "british_cuisines_df = X_y_train[X_y_train.cuisine==\"British\"]\n",
    "other_cuisines_df = X_y_train[X_y_train.cuisine!=\"British\"]\n",
    "other_cuisines_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Italian',\n",
       " 'American',\n",
       " 'Chinese',\n",
       " 'Indian',\n",
       " 'Thai and South-East Asian',\n",
       " 'Mexican',\n",
       " 'Japanese',\n",
       " 'Spanish',\n",
       " 'French',\n",
       " 'East European',\n",
       " 'Caribbean',\n",
       " 'North African',\n",
       " 'Turkish and Middle Eastern',\n",
       " 'Irish',\n",
       " 'African',\n",
       " 'Portuguese',\n",
       " 'Nordic',\n",
       " 'Pakistani',\n",
       " 'South American',\n",
       " 'Korean',\n",
       " 'Greek']"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get a list of minority cuisines\n",
    "other_cuisines = other_cuisines_df.cuisine.unique().tolist()\n",
    "other_cuisines"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "other_cuisines_upsampled = list()\n",
    "\n",
    "# Upsample the minorities\n",
    "\n",
    "for cuisine in other_cuisines:\n",
    "    cuisine_df = X_y_train[X_y_train.cuisine==cuisine]\n",
    "    cuisine_upsampled = resample(cuisine_df,\n",
    "                                 replace=True, # sample with replacement\n",
    "                                 n_samples=len(british_cuisines_df), # match number of recipes in British cuisine\n",
    "                                 random_state=1)\n",
    "    other_cuisines_upsampled.append(cuisine_upsampled)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "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>ingredients_processed</th>\n",
       "      <th>cuisine</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3347</th>\n",
       "      <td>aubergin lengthway grill oliv oil golden ball ...</td>\n",
       "      <td>Italian</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3540</th>\n",
       "      <td>white bread flour oliv oil fine salt butter bl...</td>\n",
       "      <td>Italian</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3764</th>\n",
       "      <td>globe artichok oliv oil garlic parsley doubl c...</td>\n",
       "      <td>Italian</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3638</th>\n",
       "      <td>cure salt cube lardo garlic clove cannellini b...</td>\n",
       "      <td>Italian</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3436</th>\n",
       "      <td>frozen mix berri caster sugar caster sugar egg...</td>\n",
       "      <td>Italian</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  ingredients_processed  cuisine\n",
       "3347  aubergin lengthway grill oliv oil golden ball ...  Italian\n",
       "3540  white bread flour oliv oil fine salt butter bl...  Italian\n",
       "3764  globe artichok oliv oil garlic parsley doubl c...  Italian\n",
       "3638  cure salt cube lardo garlic clove cannellini b...  Italian\n",
       "3436  frozen mix berri caster sugar caster sugar egg...  Italian"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Create a new resampled data set for minority cuisines\n",
    "other_cuisines_upsampled = pd.concat(other_cuisines_upsampled)\n",
    "other_cuisines_upsampled.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Combine the majority and the upsampled minority\n",
    "upsampled = pd.concat([british_cuisines_df, other_cuisines_upsampled])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "North African                 1099\n",
       "Spanish                       1099\n",
       "African                       1099\n",
       "Italian                       1099\n",
       "Indian                        1099\n",
       "Chinese                       1099\n",
       "British                       1099\n",
       "Pakistani                     1099\n",
       "Japanese                      1099\n",
       "Greek                         1099\n",
       "Irish                         1099\n",
       "Nordic                        1099\n",
       "East European                 1099\n",
       "Mexican                       1099\n",
       "Thai and South-East Asian     1099\n",
       "South American                1099\n",
       "Korean                        1099\n",
       "Turkish and Middle Eastern    1099\n",
       "American                      1099\n",
       "Caribbean                     1099\n",
       "Portuguese                    1099\n",
       "French                        1099\n",
       "Name: cuisine, dtype: int64"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check new class counts\n",
    "upsampled.cuisine.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "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>ingredients_processed</th>\n",
       "      <th>cuisine</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3883</th>\n",
       "      <td>oliv oil chilli garnish ginger firm firm tofu ...</td>\n",
       "      <td>Japanese</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4051</th>\n",
       "      <td>cumin chilli powder oregano bavett steak calor...</td>\n",
       "      <td>Mexican</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3098</th>\n",
       "      <td>butter digest biscuit cream chees bailey irish...</td>\n",
       "      <td>Irish</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2108</th>\n",
       "      <td>unsalt butter soften garlic bulb clove flatlea...</td>\n",
       "      <td>East European</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4253</th>\n",
       "      <td>veget oil amarillo chilli seed onion garlic cl...</td>\n",
       "      <td>South American</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3517</th>\n",
       "      <td>selfrais flour caster sugar top unsalt butter ...</td>\n",
       "      <td>Italian</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>bonein chicken thigh drumstick onion scotch bo...</td>\n",
       "      <td>African</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3848</th>\n",
       "      <td>sugar white wine vinegar chilli root ginger ga...</td>\n",
       "      <td>Japanese</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1900</th>\n",
       "      <td>veget oil chicken breast thin strip runni hone...</td>\n",
       "      <td>Chinese</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1879</th>\n",
       "      <td>greek yoghurt caster sugar lemon juic mint lea...</td>\n",
       "      <td>Caribbean</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  ingredients_processed         cuisine\n",
       "3883  oliv oil chilli garnish ginger firm firm tofu ...        Japanese\n",
       "4051  cumin chilli powder oregano bavett steak calor...         Mexican\n",
       "3098  butter digest biscuit cream chees bailey irish...           Irish\n",
       "2108  unsalt butter soften garlic bulb clove flatlea...   East European\n",
       "4253  veget oil amarillo chilli seed onion garlic cl...  South American\n",
       "3517  selfrais flour caster sugar top unsalt butter ...         Italian\n",
       "22    bonein chicken thigh drumstick onion scotch bo...         African\n",
       "3848  sugar white wine vinegar chilli root ginger ga...        Japanese\n",
       "1900  veget oil chicken breast thin strip runni hone...         Chinese\n",
       "1879  greek yoghurt caster sugar lemon juic mint lea...       Caribbean"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Overview of the upsampled dataset\n",
    "upsampled.sample(10, random_state=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature engineering using TF-IDF"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In bag of words and bag of n-grams approaches, all words in a corpus are treated equally important. TF-IDF, meanwhile, emphasizes that some words in a document are more important than others. This time, I'll try TF-IDF with unigram."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "tfidf = TfidfVectorizer()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train_new = upsampled.ingredients_processed\n",
    "y_train_new = upsampled.cuisine"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(24178, 2910) (1181, 2910)\n"
     ]
    }
   ],
   "source": [
    "# Vectorize train and test data\n",
    "X_train_transformed = tfidf.fit_transform(X_train_new)\n",
    "X_test_transformed = tfidf.transform(X_test)\n",
    "print(X_train_transformed.shape, X_test_transformed.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import Random Forest classifier\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "model = RandomForestClassifier(n_estimators=150)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Wall time: 22.2 s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(n_estimators=150)"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Train the classifier and time the training step\n",
    "%time model.fit(X_train_transformed, y_train_new)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Make class predictions for X_test_transformed\n",
    "y_predicted = model.predict(X_test_transformed)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Evaluate the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy:  0.6629974597798476\n"
     ]
    }
   ],
   "source": [
    "#Print accuracy:\n",
    "print(\"Accuracy: \", accuracy_score(y_test, y_predicted))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Function to plot confusion matrix. \n",
    "# Ref: http://scikit-learn.org/stable/auto_examples/model_selection/plot_confusion_matrix.html\n",
    "import itertools\n",
    "\n",
    "def plot_confusion_matrix(cm, classes,\n",
    "                          normalize=False,\n",
    "                          title='Confusion matrix',\n",
    "                          cmap=plt.cm.Blues):\n",
    "    \"\"\"\n",
    "    This function prints and plots the confusion matrix.\n",
    "    Normalization can be applied by setting `normalize=True`.\n",
    "    \"\"\"\n",
    "    if normalize:\n",
    "        cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "\n",
    "    plt.imshow(cm, interpolation='nearest', cmap=cmap)\n",
    "    plt.title(title)\n",
    "    plt.colorbar()\n",
    "    tick_marks = np.arange(len(classes))\n",
    "    plt.xticks(tick_marks, classes, rotation=90)\n",
    "    plt.yticks(tick_marks, classes)\n",
    "\n",
    "    fmt = '.2f' if normalize else 'd'\n",
    "    thresh = cm.max() / 2.\n",
    "    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n",
    "        plt.text(j, i, format(cm[i, j], fmt),\n",
    "                 horizontalalignment=\"center\",\n",
    "                 color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "\n",
    "    plt.tight_layout()\n",
    "    plt.ylabel('True label',fontsize=15)\n",
    "    plt.xlabel('Predicted label',fontsize=15)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x864 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# print the confusion matrix\n",
    "cnf_matrix = confusion_matrix(y_test, y_predicted)\n",
    "plt.figure(figsize=(16, 12))\n",
    "plot_confusion_matrix(cnf_matrix, classes=cuisine_list, normalize=True,\n",
    "                      title='Confusion matrix with all features - Random Forest - oversampling - unigram')\n",
    "plt.savefig(\"../static/images/cfm_rf_oversampling_uni.png\", bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "By applying the oversampling technique with unigram, the model performs slightly better to 66% overall accuracy (from nearly 65% in the RF model with oversampling, unigram and bi-grams). However, a closer look suggests that the model still completely misses out on 5 cuisines."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            precision    recall  f1-score   support\n",
      "\n",
      "                   African       0.00      0.00      0.00         7\n",
      "                  American       0.76      0.27      0.39        83\n",
      "                   British       0.60      0.89      0.72       365\n",
      "                 Caribbean       0.81      0.81      0.81        16\n",
      "                   Chinese       0.67      0.87      0.76        52\n",
      "             East European       0.67      0.25      0.36        16\n",
      "                    French       0.58      0.24      0.34       110\n",
      "                     Greek       1.00      0.09      0.17        22\n",
      "                    Indian       0.74      0.92      0.82        88\n",
      "                     Irish       0.00      0.00      0.00         5\n",
      "                   Italian       0.68      0.74      0.71       211\n",
      "                  Japanese       1.00      0.36      0.53        25\n",
      "                    Korean       0.75      0.75      0.75         4\n",
      "                   Mexican       0.75      0.73      0.74        33\n",
      "                    Nordic       1.00      0.12      0.22         8\n",
      "             North African       0.92      0.73      0.81        15\n",
      "                 Pakistani       0.00      0.00      0.00         7\n",
      "                Portuguese       0.00      0.00      0.00         3\n",
      "            South American       0.00      0.00      0.00         3\n",
      "                   Spanish       0.80      0.33      0.47        24\n",
      " Thai and South-East Asian       0.85      0.72      0.78        54\n",
      "Turkish and Middle Eastern       0.75      0.40      0.52        30\n",
      "\n",
      "                  accuracy                           0.66      1181\n",
      "                 macro avg       0.61      0.42      0.45      1181\n",
      "              weighted avg       0.67      0.66      0.63      1181\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Boss\\anaconda3\\envs\\PythonAdv\\lib\\site-packages\\sklearn\\metrics\\_classification.py:1221: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n"
     ]
    }
   ],
   "source": [
    "# Calculate classification report\n",
    "from sklearn.metrics import classification_report\n",
    "\n",
    "print(classification_report(y_test, y_predicted,\n",
    "                            target_names=cuisine_list))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda env:PythonAdv] *",
   "language": "python",
   "name": "conda-env-PythonAdv-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}