init-insights.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Social Graph Analysis Group Project\n",
    "\n",
    "### Grandma’s Cookbook, Lookbook: An Analysis of Ingredients Association With Ethnic Cuisines\n",
    "\n",
    "#### Group Members:\n",
    "* Nick Potts\n",
    "* Kristi Conry\n",
    "* Jacob Lebowitz\n",
    "* Anthony Bryson"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Possible Goals :)\n",
    "\n",
    "1) Predict popular cuisine's based on what ingediants you have <br>\n",
    "2) Essential ingrediants if you like a favorite cuisine <br>\n",
    "3) Ingrediant recommendation system ( if you lke 3 types of cuisine find top similar ingrediants ) \n",
    "\n",
    "### Just some questions for our data :)\n",
    "\n",
    "1) Which cuisine's are most similar based on ingediants? and why?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "import json\n",
    "import collections\n",
    "import re\n",
    "import pickle\n",
    "import statistics\n",
    "\n",
    "from gensim.models import word2vec\n",
    "\n",
    "# Modelling Helpers, see above the description\n",
    "from sklearn.manifold import TSNE\n",
    "\n",
    "import nltk\n",
    "from nltk.tokenize import sent_tokenize, word_tokenize\n",
    "from nltk.corpus import stopwords, wordnet\n",
    "from nltk.stem import PorterStemmer, WordNetLemmatizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv('data/cleaned_data.csv', converters={'ingredient_list_cleaned': lambda x: x[1:-1].replace(\"'\", '').split(','),\n",
    "                                                      'ingredients': lambda x: x[1:-1].replace(\"'\", '').split(',')})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "italian         7838\n",
       "mexican         6438\n",
       "southern_us     4320\n",
       "indian          3003\n",
       "chinese         2673\n",
       "french          2646\n",
       "cajun_creole    1546\n",
       "thai            1539\n",
       "japanese        1423\n",
       "greek           1175\n",
       "spanish          989\n",
       "korean           830\n",
       "vietnamese       825\n",
       "moroccan         821\n",
       "british          804\n",
       "filipino         755\n",
       "irish            667\n",
       "jamaican         526\n",
       "russian          489\n",
       "brazilian        467\n",
       "Name: cuisine, dtype: int64"
      ]
     },
     "execution_count": 93,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['cuisine'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "real_ingredients = []\n",
    "\n",
    "for ingredients in df['ingredient_list_cleaned']:\n",
    "    for ingredient in ingredients:\n",
    "        real_ingredients.append(ingredient.strip())\n",
    "        \n",
    "count_of_ingredient = collections.Counter(real_ingredients)\n",
    "count_of_ingredient.most_common(10)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "len(count_of_ingredient.keys())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1c8b3fb2b00>"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=[10,6])\n",
    "plt.xticks(rotation=70)\n",
    "sns.countplot(df['cuisine'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "count_of_ingredient_ordered = sorted(count_of_ingredient.items(), key=lambda x: x[1], reverse=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "count_of_ingredient.most_common(10)\n",
    "\n",
    "ingredients = []\n",
    "counter = []\n",
    "\n",
    "for counted_ingredients in count_of_ingredient.most_common(40):\n",
    "    ingredients.append(counted_ingredients[0])\n",
    "    counter.append(counted_ingredients[1])\n",
    "    \n",
    "fig = plt.figure(figsize=[16,8])\n",
    "\n",
    "plt.xticks(rotation=70)\n",
    "\n",
    "sns.barplot(x=ingredients, y=counter)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sentences = []\n",
    "\n",
    "for ingredients in df['ingredient_list_cleaned']:\n",
    "    cleaned_ing = []\n",
    "    for ingredient in ingredients:\n",
    "        ingredient = re.sub('r\\(.*oz\\)|minced|poweder|chopped|sliced|paste|crushed|ground|diced', '', ingredient)\n",
    "        cleaned_ing.append(ingredient.strip())\n",
    "    sentences.append(cleaned_ing)\n",
    "    \n",
    "print(len(sentences))\n",
    "print(sentences[:10])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "keys = count_of_ingredient.keys()\n",
    "values = count_of_ingredient.values()\n",
    "\n",
    "fig1, ax1 = plt.subplots()\n",
    "ax1.set_title('Number of Ingredients')\n",
    "ax1.boxplot(list(values), showfliers=False)\n",
    "\n",
    "len_of_sentences = df['ingredient_list_cleaned'].apply(lambda x: len(x))\n",
    "print(type(len_of_sentences))\n",
    "print(len_of_sentences.describe())\n",
    "\n",
    "fig2, ax2 = plt.subplots()\n",
    "ax2.set_title('Length of Sentences')\n",
    "ax2.boxplot(len_of_sentences, showfliers=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "num_of_features = 250 # Dimensionality of the word vectors.\n",
    "\n",
    "min_word_count = 4 # Ignores all words with total frequency lower than this.\n",
    "\n",
    "num_workers = 4 # Use these many worker threads to train the model\n",
    "\n",
    "window_context = 6 # Maximum distance between the current and predicted word within a sentence.\n",
    "\n",
    "sample = 1e-3 # The threshold for configuring which higher-frequency words are randomly downsampled."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = word2vec.Word2Vec(sentences, size=num_of_features,\n",
    "                          min_count=min_word_count, workers=num_workers,\n",
    "                          window=window_context, sample=sample)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.wv.most_similar('jalapeno chili')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.wv.similarity('fresh pineappl', 'milk')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def display_closestwords_tsnescatterplot(model, word):\n",
    "    \n",
    "    arr = np.empty((0,250), dtype='f')\n",
    "    word_labels = [word]\n",
    "\n",
    "    # get close words\n",
    "    close_words = model.similar_by_word(word)\n",
    "    \n",
    "    # add the vector for each of the closest words to the array\n",
    "    arr = np.append(arr, np.array([model[word]]), axis=0)\n",
    "    for wrd_score in close_words:\n",
    "        wrd_vector = model[wrd_score[0]]\n",
    "        word_labels.append(wrd_score[0])\n",
    "        arr = np.append(arr, np.array([wrd_vector]), axis=0)\n",
    "        \n",
    "    # find tsne coords for 2 dimensions\n",
    "    tsne = TSNE(n_components=2, random_state=0)\n",
    "    np.set_printoptions(suppress=True)\n",
    "    Y = tsne.fit_transform(arr)\n",
    "\n",
    "    x_coords = Y[:, 0]\n",
    "    y_coords = Y[:, 1]\n",
    "    # display scatter plot\n",
    "    \n",
    "    plt.scatter(x_coords, y_coords)\n",
    "\n",
    "    for label, x, y in zip(word_labels, x_coords, y_coords):\n",
    "        plt.annotate(label, xy=(x, y), xytext=(0, 0), textcoords='offset points')\n",
    "    plt.xlim(x_coords.min()+0.00005, x_coords.max()+0.00005)\n",
    "    plt.ylim(y_coords.min()+0.00005, y_coords.max()+0.00005)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('orang bell pepper', 0.9075711369514465),\n",
       " ('cherri tomato', 0.8766277432441711),\n",
       " ('avocado', 0.8739045858383179),\n",
       " ('yellow bell pepper', 0.8732463121414185),\n",
       " ('grape tomato', 0.8645813465118408),\n",
       " ('roma tomato', 0.8590236902236938),\n",
       " ('reducedfat sour cream', 0.8582755327224731),\n",
       " ('bake tortilla chip', 0.854686439037323),\n",
       " ('reduc sodium black bean', 0.8538311123847961),\n",
       " ('jicama', 0.8485833406448364)]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.wv.most_similar('romain lettuc')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Library/Frameworks/Python.framework/Versions/3.7/lib/python3.7/site-packages/ipykernel_launcher.py:7: DeprecationWarning: Call to deprecated `similar_by_word` (Method will be removed in 4.0.0, use self.wv.similar_by_word() instead).\n",
      "  import sys\n",
      "/Library/Frameworks/Python.framework/Versions/3.7/lib/python3.7/site-packages/ipykernel_launcher.py:10: DeprecationWarning: Call to deprecated `__getitem__` (Method will be removed in 4.0.0, use self.wv.__getitem__() instead).\n",
      "  # Remove the CWD from sys.path while we load stuff.\n",
      "/Library/Frameworks/Python.framework/Versions/3.7/lib/python3.7/site-packages/ipykernel_launcher.py:12: DeprecationWarning: Call to deprecated `__getitem__` (Method will be removed in 4.0.0, use self.wv.__getitem__() instead).\n",
      "  if sys.path[0] == '':\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "display_closestwords_tsnescatterplot(model, 'romain lettuc')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Library/Frameworks/Python.framework/Versions/3.7/lib/python3.7/site-packages/ipykernel_launcher.py:6: DeprecationWarning: Call to deprecated `__getitem__` (Method will be removed in 4.0.0, use self.wv.__getitem__() instead).\n",
      "  \n"
     ]
    }
   ],
   "source": [
    "    \"Creates and TSNE model and plots it\"\n",
    "    labels = []\n",
    "    tokens = []\n",
    "\n",
    "    for word in model.wv.vocab:\n",
    "        tokens.append(model[word])\n",
    "        labels.append(word)\n",
    "        \n",
    "    #print(tokens[:5])\n",
    "    #print(labels[:5])\n",
    "    \n",
    "    tsne_model = TSNE(perplexity=40, n_components=2, init='pca', n_iter=2500, random_state=23)\n",
    "    '''\n",
    "    tool to visualize high-dimensional data. It converts similarities between data points to joint\n",
    "    probabilities and tries to minimize the Kullback-Leibler divergence between the joint probabilities \n",
    "    of the low-dimensional embedding and the high-dimensional data\n",
    "    '''\n",
    "    new_values = tsne_model.fit_transform(tokens)\n",
    "    #print(new_values)\n",
    "    #print(new_values[:5])\n",
    "    \n",
    "    x = []\n",
    "    y = []\n",
    "    for value in new_values:\n",
    "        x.append(value[0])\n",
    "        y.append(value[1])\n",
    "    \n",
    "    \n",
    "    #print(x[0])\n",
    "    #print(y[0])\n",
    "    '''\n",
    "    plt.figure(figsize=(16, 16)) \n",
    "    for i in range(len(x)):\n",
    "        plt.scatter(x[i],y[i])\n",
    "        #plt.annotate('',\n",
    "         #            xy=(x[i], y[i]),\n",
    "          #           xytext=(5, 2),\n",
    "           #          textcoords='offset points',\n",
    "            #         ha='right',\n",
    "             #        va='bottom')\n",
    "    plt.show()\n",
    "    '''"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "classifier = KNeighborsClassifier(n_neighbors=5)\n",
    "classifier.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.wv.most_similar('romain lettuc')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 1) User Search for Ingredient that they enjoy --> Once Found reeccomend ingredients"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from gensim.models import Word2Vec, KeyedVectors   \n",
    "model.wv.save_word2vec_format('model.bin', binary=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.save(\"model.model\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "ingred = []\n",
    "for key, value in count_of_ingredient.items():\n",
    "    if value > 4:\n",
    "        if key != '':\n",
    "            if key[0] == ' ':\n",
    "                key = key[1:]\n",
    "            ingred.append(key)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3328"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(ingred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "cuisineObj = []\n",
    "for key, value in cuisine_counter.items():\n",
    "    cuisineObj.append(key)\n",
    "    \n",
    "#print(len(cuisineObj))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('ingred_list_v2.pkl', 'wb') as f:\n",
    "    pickle.dump(ingred, f)\n",
    "    \n",
    "    \n",
    "with open('cuisine_obj.pkl', 'wb') as f:\n",
    "    pickle.dump(cuisineObj, f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['greek',\n",
       " 'southern_us',\n",
       " 'filipino',\n",
       " 'indian',\n",
       " 'jamaican',\n",
       " 'spanish',\n",
       " 'italian',\n",
       " 'mexican',\n",
       " 'chinese',\n",
       " 'british',\n",
       " 'thai',\n",
       " 'vietnamese',\n",
       " 'cajun_creole',\n",
       " 'brazilian',\n",
       " 'french',\n",
       " 'japanese',\n",
       " 'irish',\n",
       " 'korean',\n",
       " 'moroccan',\n",
       " 'russian']"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cuisineObj"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0              [greek]\n",
       "1        [southern_us]\n",
       "2           [filipino]\n",
       "3             [indian]\n",
       "4             [indian]\n",
       "             ...      \n",
       "39769          [irish]\n",
       "39770        [italian]\n",
       "39771          [irish]\n",
       "39772        [chinese]\n",
       "39773        [mexican]\n",
       "Name: cuisine, Length: 39774, dtype: object"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['cuisine']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "ename": "IndentationError",
     "evalue": "unexpected indent (<ipython-input-15-5464665df482>, line 5)",
     "output_type": "error",
     "traceback": [
      "\u001b[0;36m  File \u001b[0;32m\"<ipython-input-15-5464665df482>\"\u001b[0;36m, line \u001b[0;32m5\u001b[0m\n\u001b[0;31m    if ingred is not in dict\u001b[0m\n\u001b[0m    ^\u001b[0m\n\u001b[0;31mIndentationError\u001b[0m\u001b[0;31m:\u001b[0m unexpected indent\n"
     ]
    }
   ],
   "source": [
    "dict = {}\n",
    "\n",
    "# iterate over each row\n",
    "    # iterate over ingredient list cleaned\n",
    "        if ingred is not in dict\n",
    "            add key to dict withvalue in list\n",
    "        else\n",
    "            append cusine to value lsit at dict[key]\n",
    "            "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['cuisine'] = df['cuisine'].apply(lambda x: x[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "cook_book = {}\n",
    "\n",
    "for index, row in df.iterrows():\n",
    "    \n",
    "    #print(row['cuisine'], row['ingredient_list_cleaned'])\n",
    "\n",
    "    #print(row['cuisine'])\n",
    "    for ingred in row['ingredient_list_cleaned']:\n",
    "        if ingred[0] == ' ':\n",
    "            ingred = ingred[1:]\n",
    "        #print(row['cuisine'])\n",
    "        if ingred not in cook_book.keys():\n",
    "            #print(ingred)\n",
    "            cook_book[ingred] = dict()\n",
    "            cook_book[ingred][row['cuisine']] = 1\n",
    "            \n",
    "        else:\n",
    "            if row['cuisine'] in cook_book[ingred].keys():\n",
    "                cook_book[ingred][row['cuisine']] += 1\n",
    "            else:\n",
    "                cook_book[ingred][row['cuisine']] = 1\n",
    "                \n",
    "        cook_book[ingred] = dict(sorted(cook_book[ingred].items(), key=lambda kv: kv[1], reverse=True))\n",
    "        \n",
    "            #cook_book[ingred][row['cuisine']] += 1\n",
    "            \n",
    "        \n",
    "            \n",
    "    \n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('cook_book_v3.pkl', 'wb') as f:\n",
    "    pickle.dump(cook_book, f)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "cook_book['grill chicken breast']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "cook_book"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}