Cooking-4.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn.naive_bayes import MultinomialNB\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.model_selection import train_test_split\n",
    "from time import time\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn import tree\n",
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "import json\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "from sklearn.feature_extraction.text import TfidfTransformer\n",
    "from sklearn.feature_selection import chi2\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.svm import LinearSVC\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.model_selection import learning_curve\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'dict'>\n",
      "{'id': 10259, 'cuisine': 'greek', 'ingredients': ['romaine lettuce', 'black olives', 'grape tomatoes', 'garlic', 'pepper', 'purple onion', 'seasoning', 'garbanzo beans', 'feta cheese crumbles']}\n",
      "\n",
      "Number of recipes:  39774 \n",
      "\n",
      "['romaine lettuce', 'black olives', 'grape tomatoes', 'garlic', 'pepper', 'purple onion', 'seasoning', 'garbanzo beans', 'feta cheese crumbles']\n",
      "9\n",
      "65\n",
      "3885\n",
      "39774\n"
     ]
    }
   ],
   "source": [
    "with open(\"train.json\") as f:\n",
    "    datastore = json.loads(f.read())\n",
    "f.close()\n",
    "with open(\"test.json\") as f:\n",
    "    test = json.loads(f.read())\n",
    "f.close()\n",
    "print(type(datastore[0]))\n",
    "print(datastore[0])\n",
    "print(\"\\nNumber of recipes: \",len(datastore),\"\\n\")\n",
    "print(datastore[0]['ingredients'])\n",
    "print(len(datastore[0]['ingredients']))\n",
    "length = len(datastore)\n",
    "length1 = len(test)\n",
    "max = 0\n",
    "pos = 0\n",
    "for x in range(length):\n",
    "    if(len(datastore[x]['ingredients']) > max):\n",
    "        max = len(datastore[x]['ingredients'])\n",
    "        pos = x\n",
    "print(max)\n",
    "print(datastore[pos]['id'])\n",
    "print(length)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "5000\n",
      "10000\n",
      "15000\n",
      "20000\n",
      "25000\n",
      "30000\n",
      "35000\n",
      "greek\n",
      "romaine lettuce,black olives,grape tomatoes,garlic,pepper,purple onion,seasoning,garbanzo beans,feta cheese crumbles\n"
     ]
    }
   ],
   "source": [
    "cuisine = []\n",
    "ingre = []\n",
    "full = []\n",
    "for i in range(length):\n",
    "    cuisine.append(datastore[i]['cuisine'])\n",
    "    for j in range(len(datastore[i]['ingredients'])):\n",
    "        ingre.append(datastore[i]['ingredients'][j])\n",
    "    li = [y for y in ingre]\n",
    "    li = \",\".join(map(str, li))\n",
    "    full.append(li)\n",
    "    ingre = []\n",
    "    #if(i == 5000):\n",
    "    #    break\n",
    "    if(i%5000 == 0):\n",
    "        print(i)\n",
    "df = pd.DataFrame({'cuisine':cuisine,'ingredients':full})\n",
    "for i in df.loc[0]:\n",
    "    print(i)\n",
    "#print(df.head())\n",
    "#df.groupby('cuisine').cuisine.count().plot.bar(ylim=0)\n",
    "#plt.show()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "       cuisine                                        ingredients\n",
      "0        greek  romaine lettuce,black olives,grape tomatoes,ga...\n",
      "1  southern_us  plain flour,ground pepper,salt,tomatoes,ground...\n",
      "2     filipino  eggs,pepper,salt,mayonaise,cooking oil,green c...\n",
      "3       indian                     water,vegetable oil,wheat,salt\n",
      "4       indian  black pepper,shallots,cornflour,cayenne pepper...\n",
      "['brazilian' 'british' 'cajun_creole' 'chinese' 'filipino' 'french'\n",
      " 'greek' 'indian' 'irish' 'italian' 'jamaican' 'japanese' 'korean'\n",
      " 'mexican' 'moroccan' 'russian' 'southern_us' 'spanish' 'thai'\n",
      " 'vietnamese']\n",
      "brazilian\n",
      "{'brazilian': 0, 'british': 1, 'cajun_creole': 2, 'chinese': 3, 'filipino': 4, 'french': 5, 'greek': 6, 'indian': 7, 'irish': 8, 'italian': 9, 'jamaican': 10, 'japanese': 11, 'korean': 12, 'mexican': 13, 'moroccan': 14, 'russian': 15, 'southern_us': 16, 'spanish': 17, 'thai': 18, 'vietnamese': 19}\n",
      "   cuisine                                        ingredients\n",
      "0        6  romaine lettuce,black olives,grape tomatoes,ga...\n",
      "1       16  plain flour,ground pepper,salt,tomatoes,ground...\n",
      "2        4  eggs,pepper,salt,mayonaise,cooking oil,green c...\n",
      "3        7                     water,vegetable oil,wheat,salt\n",
      "4        7  black pepper,shallots,cornflour,cayenne pepper...\n"
     ]
    }
   ],
   "source": [
    "print(df.head())\n",
    "df1 = df.copy()\n",
    "cuis = np.unique(df1['cuisine'].values)\n",
    "print(cuis)\n",
    "print(cuis[0])\n",
    "b = {cuis[i]:i for i in range(len(cuis))}\n",
    "print(b)\n",
    "df1 = df1.replace({'cuisine':b})\n",
    "print(df1.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "romaine lettuce,black olives,grape tomatoes,garlic,pepper,purple onion,seasoning,garbanzo beans,feta cheese crumbles\n",
      "(39774, 17927)\n"
     ]
    }
   ],
   "source": [
    "tfidf = TfidfVectorizer(sublinear_tf=True, min_df=5,norm='l2',\\\n",
    "                       encoding='latin-1',ngram_range=(1,2))\n",
    "print(df1['ingredients'][0])\n",
    "features = tfidf.fit_transform(df1.ingredients).toarray()\n",
    "labels = df1['cuisine']\n",
    "print(features.shape)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0    romaine lettuce,black olives,grape tomatoes,ga...\n",
      "1    plain flour,ground pepper,salt,tomatoes,ground...\n",
      "2    eggs,pepper,salt,mayonaise,cooking oil,green c...\n",
      "3                       water,vegetable oil,wheat,salt\n",
      "4    black pepper,shallots,cornflour,cayenne pepper...\n",
      "Name: ingredients, dtype: object\n",
      "0     6\n",
      "1    16\n",
      "2     4\n",
      "3     7\n",
      "4     7\n",
      "Name: cuisine, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "print(df1['ingredients'].head())\n",
    "print(df1['cuisine'].head())\n",
    "X_train, X_test, y_train, y_test =\\\n",
    "train_test_split(df1['ingredients'],\\\n",
    "                 df1['cuisine'], random_state = 0)\n",
    "count_vect = CountVectorizer()\n",
    "X_train_counts = count_vect.fit_transform(X_train)\n",
    "tfidf_transformer = TfidfTransformer()\n",
    "X_train_tfidf = tfidf_transformer.fit_transform(X_train_counts)\n",
    "X_train_tfidf1 = tfidf.fit_transform(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5110     ground black pepper,pizza doughs,olive oil,gar...\n",
      "26915    olive oil,onions,lime juice,salt,chopped cilan...\n",
      "16284    white onion,whole milk,taco seasoning,ground b...\n",
      "9538     vanilla beans,red wine,sugar,large egg yolks,a...\n",
      "4243     pepper,butter,grated parmesan cheese,salt,minc...\n",
      "Name: ingredients, dtype: object\n",
      "  (0, 2224)\t1\n",
      "  (0, 849)\t1\n",
      "  (0, 1082)\t1\n",
      "  (0, 489)\t2\n",
      "  (0, 1841)\t1\n",
      "  (0, 1123)\t1\n",
      "  (0, 1909)\t1\n",
      "  (0, 1044)\t1\n",
      "  (0, 1758)\t1\n",
      "  (0, 1765)\t1\n",
      "  (0, 806)\t1\n",
      "  (0, 1968)\t1\n",
      "  (0, 1891)\t1\n",
      "  (0, 232)\t1\n",
      "  (0, 1147)\t1\n",
      "  (1, 723)\t1\n",
      "  (1, 190)\t1\n",
      "  (1, 2121)\t1\n",
      "  (1, 2631)\t1\n",
      "  (1, 494)\t1\n",
      "  (1, 1175)\t1\n",
      "  (1, 308)\t1\n",
      "  (1, 503)\t1\n",
      "  (1, 2361)\t1\n",
      "  (1, 266)\t1\n",
      "  :\t:\n",
      "  (29828, 1854)\t1\n",
      "  (29828, 1491)\t1\n",
      "  (29828, 1668)\t1\n",
      "  (29828, 1009)\t1\n",
      "  (29828, 1283)\t1\n",
      "  (29828, 2394)\t1\n",
      "  (29828, 2340)\t1\n",
      "  (29828, 2247)\t1\n",
      "  (29828, 181)\t1\n",
      "  (29828, 2276)\t1\n",
      "  (29828, 542)\t1\n",
      "  (29828, 489)\t1\n",
      "  (29828, 1758)\t1\n",
      "  (29828, 1765)\t1\n",
      "  (29828, 1968)\t1\n",
      "  (29828, 1147)\t1\n",
      "  (29829, 1352)\t1\n",
      "  (29829, 130)\t1\n",
      "  (29829, 355)\t1\n",
      "  (29829, 1009)\t1\n",
      "  (29829, 964)\t1\n",
      "  (29829, 2068)\t1\n",
      "  (29829, 45)\t1\n",
      "  (29829, 639)\t1\n",
      "  (29829, 2794)\t1\n",
      "  (0, 1147)\t0.13110767855831948\n",
      "  (0, 232)\t0.14706040131648623\n",
      "  (0, 1891)\t0.10695762632137122\n",
      "  (0, 1968)\t0.3611320291154489\n",
      "  (0, 806)\t0.4252049643077899\n",
      "  (0, 1765)\t0.14318307141200787\n",
      "  (0, 1758)\t0.10130357435086648\n",
      "  (0, 1044)\t0.11051808469697873\n",
      "  (0, 1909)\t0.38116136710643944\n",
      "  (0, 1123)\t0.21915431315370412\n",
      "  (0, 1841)\t0.22432322425668189\n",
      "  (0, 489)\t0.3050688686164524\n",
      "  (0, 1082)\t0.37502673644460544\n",
      "  (0, 849)\t0.33062405531214556\n",
      "  (0, 2224)\t0.09430048030909556\n",
      "  (1, 1147)\t0.13648403790960037\n",
      "  (1, 232)\t0.15309093722799896\n",
      "  (1, 1891)\t0.22268731909662173\n",
      "  (1, 1765)\t0.1490546088642452\n",
      "  (1, 1758)\t0.10545775071390223\n",
      "  (1, 2224)\t0.09816747936449081\n",
      "  (1, 1771)\t0.14400578038976675\n",
      "  (1, 1458)\t0.20905661111166385\n",
      "  (1, 1318)\t0.1796019850027954\n",
      "  (1, 542)\t0.18014492388558934\n",
      "  :\t:\n",
      "  (29828, 542)\t0.1418867000811945\n",
      "  (29828, 2276)\t0.19742904566944666\n",
      "  (29828, 181)\t0.18681462981744568\n",
      "  (29828, 2247)\t0.11946783130528618\n",
      "  (29828, 2340)\t0.19191356853856187\n",
      "  (29828, 2394)\t0.19904044138727883\n",
      "  (29828, 1283)\t0.22398066768919336\n",
      "  (29828, 1009)\t0.2179267921175214\n",
      "  (29828, 1668)\t0.22045410935955156\n",
      "  (29828, 1491)\t0.19938796125092942\n",
      "  (29828, 1854)\t0.1860774309531982\n",
      "  (29828, 2630)\t0.3845962581245581\n",
      "  (29828, 1431)\t0.26769324341572526\n",
      "  (29828, 2437)\t0.22226133229800937\n",
      "  (29828, 805)\t0.2887862613438995\n",
      "  (29828, 2129)\t0.3266550341542504\n",
      "  (29829, 2794)\t0.32324763915415605\n",
      "  (29829, 639)\t0.2519504359527668\n",
      "  (29829, 45)\t0.2528422928476206\n",
      "  (29829, 2068)\t0.25295446514566955\n",
      "  (29829, 964)\t0.20968295089107683\n",
      "  (29829, 1009)\t0.34909769157144577\n",
      "  (29829, 355)\t0.3842155461321142\n",
      "  (29829, 130)\t0.3683112972144534\n",
      "  (29829, 1352)\t0.5049810089643049\n"
     ]
    }
   ],
   "source": [
    "print((X_train.head()))\n",
    "print(X_train_counts)\n",
    "print(X_train_tfidf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "clf = MultinomialNB().fit(X_train_tfidf, y_train)\n",
    "lgr = LogisticRegression(random_state=0).fit(X_train_tfidf, y_train)\n",
    "lsvc = LinearSVC().fit(X_train_tfidf,y_train)\n",
    "#lsvc1 = LinearSVC().fit(X_train_tfidf1,y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['plain flour', 'ground pepper', 'salt', 'tomatoes', 'ground black pepper', 'thyme', 'eggs', 'green tomatoes', 'yellow corn meal', 'milk', 'vegetable oil']\n",
      "plain flour,ground pepper,salt,tomatoes,ground black pepper,thyme,eggs,green tomatoes,yellow corn meal,milk,vegetable oil\n",
      "['southern_us']\n"
     ]
    }
   ],
   "source": [
    "str1 = datastore[1]['ingredients']\n",
    "print(str1)\n",
    "str1 = \",\".join(map(str, str1))\n",
    "print(str1)\n",
    "print( cuis[(clf.predict(count_vect.transform([str1])))] )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# <!-- collapse=True -->\n",
    "\n",
    "# Modified from http://scikit-learn.org/stable/auto_examples/plot_learning_curve.html\n",
    "from sklearn.learning_curve import learning_curve\n",
    "def plot_learning_curve(estimator, title, X, y, ylim=None, cv=None,\n",
    "                        train_sizes=np.linspace(.1, 1.0, 5)):\n",
    "    \"\"\"\n",
    "    Generate a simple plot of the test and traning learning curve.\n",
    "\n",
    "    Parameters\n",
    "    ----------\n",
    "    estimator : object type that implements the \"fit\" and \"predict\" methods\n",
    "        An object of that type which is cloned for each validation.\n",
    "\n",
    "    title : string\n",
    "        Title for the chart.\n",
    "\n",
    "    X : array-like, shape (n_samples, n_features)\n",
    "        Training vector, where n_samples is the number of samples and\n",
    "        n_features is the number of features.\n",
    "\n",
    "    y : array-like, shape (n_samples) or (n_samples, n_features), optional\n",
    "        Target relative to X for classification or regression;\n",
    "        None for unsupervised learning.\n",
    "\n",
    "    ylim : tuple, shape (ymin, ymax), optional\n",
    "        Defines minimum and maximum yvalues plotted.\n",
    "\n",
    "    cv : integer, cross-validation generator, optional\n",
    "        If an integer is passed, it is the number of folds (defaults to 3).\n",
    "        Specific cross-validation objects can be passed, see\n",
    "        sklearn.cross_validation module for the list of possible objects\n",
    "    \"\"\"\n",
    "    \n",
    "    plt.figure()\n",
    "    train_sizes, train_scores, test_scores = learning_curve(\n",
    "        estimator, X, y, cv=5, n_jobs=1, train_sizes=train_sizes)\n",
    "    train_scores_mean = np.mean(train_scores, axis=1)\n",
    "    train_scores_std = np.std(train_scores, axis=1)\n",
    "    test_scores_mean = np.mean(test_scores, axis=1)\n",
    "    test_scores_std = np.std(test_scores, axis=1)\n",
    "\n",
    "    plt.fill_between(train_sizes, train_scores_mean - train_scores_std,\n",
    "                     train_scores_mean + train_scores_std, alpha=0.1,\n",
    "                     color=\"r\")\n",
    "    plt.fill_between(train_sizes, test_scores_mean - test_scores_std,\n",
    "                     test_scores_mean + test_scores_std, alpha=0.1, color=\"g\")\n",
    "    plt.plot(train_sizes, train_scores_mean, 'o-', color=\"r\",\n",
    "             label=\"Training score\")\n",
    "    plt.plot(train_sizes, test_scores_mean, 'o-', color=\"g\",\n",
    "             label=\"Cross-validation score\")\n",
    "\n",
    "    plt.xlabel(\"Training examples\")\n",
    "    plt.ylabel(\"Score\")\n",
    "    plt.legend(loc=\"best\")\n",
    "    plt.grid(\"on\") \n",
    "    if ylim:\n",
    "        plt.ylim(ylim)\n",
    "    plt.title(title)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAEWCAYAAACXGLsWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ne8klToi93w0AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fcfb2c3c080>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_learning_curve(LinearSVC(), \n",
    "                    \"LinearSVC(C=0.1, penalty='l1')\", \n",
    "                    features, labels,\n",
    "                    train_sizes=np.linspace(.05, 0.2, 5))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',\n",
      "            max_depth=3, max_features='auto', max_leaf_nodes=None,\n",
      "            min_impurity_decrease=0.0, min_impurity_split=None,\n",
      "            min_samples_leaf=1, min_samples_split=2,\n",
      "            min_weight_fraction_leaf=0.0, n_estimators=200, n_jobs=1,\n",
      "            oob_score=False, random_state=0, verbose=0, warm_start=False) : [0.32921128 0.32649263]\n",
      "\n",
      "\n",
      "\n",
      "LinearSVC(C=1.0, class_weight=None, dual=True, fit_intercept=True,\n",
      "     intercept_scaling=1, loss='squared_hinge', max_iter=1000,\n",
      "     multi_class='ovr', penalty='l2', random_state=None, tol=0.0001,\n",
      "     verbose=0) : [0.77771075 0.77023289]\n",
      "\n",
      "\n",
      "\n",
      "MultinomialNB(alpha=1.0, class_prior=None, fit_prior=True) : [0.60267431 0.60696142]\n",
      "\n",
      "\n",
      "\n",
      "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
      "          intercept_scaling=1, max_iter=100, multi_class='ovr', n_jobs=1,\n",
      "          penalty='l2', random_state=0, solver='liblinear', tol=0.0001,\n",
      "          verbose=0, warm_start=False) : [0.75202332 0.74739701]\n",
      "\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "CV = 1\n",
    "models = [\n",
    "    RandomForestClassifier(n_estimators=200, max_depth=3, random_state=0),\n",
    "    LinearSVC(),\n",
    "    MultinomialNB(),\n",
    "    LogisticRegression(random_state=0),\n",
    "]\n",
    "\n",
    "for model in models:\n",
    "    accuracy = cross_val_score(model,features,labels,\\\n",
    "                            scoring='accuracy',cv=2)\n",
    "    print(str(model),\":\",accuracy)\n",
    "    print(\"\\n\\n\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "5000\n",
      "      id                                        ingredients\n",
      "0  18009  baking powder,eggs,all-purpose flour,raisins,m...\n",
      "1  28583  sugar,egg yolks,corn starch,cream of tartar,ba...\n",
      "2  41580  sausage links,fennel bulb,fronds,olive oil,cub...\n",
      "3  29752  meat cuts,file powder,smoked sausage,okra,shri...\n",
      "4  35687  ground black pepper,salt,sausage casings,leeks...\n",
      "baking powder,eggs,all-purpose flour,raisins,milk,white sugar\n",
      "\n",
      "\n",
      "\n",
      "['baking powder,eggs,all-purpose flour,raisins,milk,white sugar', 'sugar,egg yolks,corn starch,cream of tartar,bananas,vanilla wafers,milk,vanilla extract,toasted pecans,egg whites,light rum']\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "          cuisine     id\n",
      "0   [southern_us]  18009\n",
      "1   [southern_us]  28583\n",
      "2       [italian]  41580\n",
      "3  [cajun_creole]  29752\n",
      "4       [italian]  35687\n",
      "        cuisine     id\n",
      "0   southern_us  18009\n",
      "1   southern_us  28583\n",
      "2       italian  41580\n",
      "3  cajun_creole  29752\n",
      "4       italian  35687\n"
     ]
    }
   ],
   "source": [
    "id1 = []\n",
    "ingre1 = []\n",
    "full1 = []\n",
    "for i in range(length1):\n",
    "    id1.append(test[i]['id'])\n",
    "    for j in range(len\n",
    "                   (test[i]['ingredients'])):\n",
    "        ingre1.append(test[i]['ingredients'][j])\n",
    "    li1 = [y for y in ingre1]\n",
    "    li1 = \",\".join(map(str, li1))\n",
    "    full1.append(li1)\n",
    "    ingre1 = []\n",
    "    #if(i == 5000):\n",
    "    #    break\n",
    "    if(i%5000 == 0):\n",
    "        print(i)\n",
    "        \n",
    "full2 = []\n",
    "for k in range(len(full1)):\n",
    "    full2.append(full1[k])\n",
    "\n",
    "df2 = pd.DataFrame({'id':id1,'ingredients':full1})\n",
    "print(df2.head())\n",
    "print(df2['ingredients'][0])\n",
    "ans = []\n",
    "\n",
    "print(\"\\n\\n\")\n",
    "print(full2[:2])\n",
    "print(\"\\n\\n\\n\")\n",
    "\n",
    "for k in range(len(full1)):\n",
    "    ans.append((cuis[(\\\n",
    "    clf.predict(count_vect.transform([full1[k]])))]))\n",
    "\n",
    "out = pd.DataFrame({'id':df2['id'], 'cuisine':ans})\n",
    "'''df2['ingredients'] = \",\".join(map(str, df2['ingredients']))\n",
    "print(df2.head())\n",
    "ans = pd.DataFrame(columns=['cuisine'])\n",
    "k = 0\n",
    "print(df2['ingredients'][0])\n",
    "print(\"\\t\\t\\tTesting\")'''\n",
    "'''for i in df2['ingredients']:\n",
    "    temp = pd.DataFrame(columns=['cuisine'])\n",
    "    temp['cuisine'] = (cuis[(\\\n",
    "    clf.predict(count_vect.transform([i])))])\n",
    "    #print(i)\n",
    "    print(temp)\n",
    "    ans = ans.append(temp)\n",
    "    k+=10\n",
    "    if(k == 1):\n",
    "        break'''\n",
    "#print(ans)\n",
    "print(out.head())\n",
    "out['cuisine'] = out['cuisine'].str[0]\n",
    "print(out.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "file = \"submission.csv\"\n",
    "out.to_csv(file,index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "5000\n",
      "      id                                        ingredients\n",
      "0  18009  baking powder,eggs,all-purpose flour,raisins,m...\n",
      "1  28583  sugar,egg yolks,corn starch,cream of tartar,ba...\n",
      "2  41580  sausage links,fennel bulb,fronds,olive oil,cub...\n",
      "3  29752  meat cuts,file powder,smoked sausage,okra,shri...\n",
      "4  35687  ground black pepper,salt,sausage casings,leeks...\n",
      "baking powder,eggs,all-purpose flour,raisins,milk,white sugar\n",
      "\n",
      "\n",
      "\n",
      "['baking powder,eggs,all-purpose flour,raisins,milk,white sugar', 'sugar,egg yolks,corn starch,cream of tartar,bananas,vanilla wafers,milk,vanilla extract,toasted pecans,egg whites,light rum']\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "          cuisine     id\n",
      "0       [british]  18009\n",
      "1   [southern_us]  28583\n",
      "2       [italian]  41580\n",
      "3  [cajun_creole]  29752\n",
      "4       [italian]  35687\n",
      "        cuisine     id\n",
      "0       british  18009\n",
      "1   southern_us  28583\n",
      "2       italian  41580\n",
      "3  cajun_creole  29752\n",
      "4       italian  35687\n"
     ]
    }
   ],
   "source": [
    "id1 = []\n",
    "ingre1 = []\n",
    "full1 = []\n",
    "for i in range(length1):\n",
    "    id1.append(test[i]['id'])\n",
    "    for j in range(len\n",
    "                   (test[i]['ingredients'])):\n",
    "        ingre1.append(test[i]['ingredients'][j])\n",
    "    li1 = [y for y in ingre1]\n",
    "    li1 = \",\".join(map(str, li1))\n",
    "    full1.append(li1)\n",
    "    ingre1 = []\n",
    "    #if(i == 5000):\n",
    "    #    break\n",
    "    if(i%5000 == 0):\n",
    "        print(i)\n",
    "        \n",
    "full2 = []\n",
    "for k in range(len(full1)):\n",
    "    full2.append(full1[k])\n",
    "\n",
    "df2 = pd.DataFrame({'id':id1,'ingredients':full1})\n",
    "print(df2.head())\n",
    "print(df2['ingredients'][0])\n",
    "ans = []\n",
    "\n",
    "print(\"\\n\\n\")\n",
    "print(full2[:2])\n",
    "print(\"\\n\\n\\n\")\n",
    "\n",
    "for k in range(len(full1)):\n",
    "    ans.append((cuis[(\\\n",
    "    lgr.predict(count_vect.transform([full1[k]])))]))\n",
    "\n",
    "out = pd.DataFrame({'id':df2['id'], 'cuisine':ans})\n",
    "'''df2['ingredients'] = \",\".join(map(str, df2['ingredients']))\n",
    "print(df2.head())\n",
    "ans = pd.DataFrame(columns=['cuisine'])\n",
    "k = 0\n",
    "print(df2['ingredients'][0])\n",
    "print(\"\\t\\t\\tTesting\")'''\n",
    "'''for i in df2['ingredients']:\n",
    "    temp = pd.DataFrame(columns=['cuisine'])\n",
    "    temp['cuisine'] = (cuis[(\\\n",
    "    clf.predict(count_vect.transform([i])))])\n",
    "    #print(i)\n",
    "    print(temp)\n",
    "    ans = ans.append(temp)\n",
    "    k+=10\n",
    "    if(k == 1):\n",
    "        break'''\n",
    "#print(ans)\n",
    "print(out.head())\n",
    "out['cuisine'] = out['cuisine'].str[0]\n",
    "print(out.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "file = \"submission2.csv\"\n",
    "out.to_csv(file,index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "5000\n",
      "      id                                        ingredients\n",
      "0  18009  baking powder,eggs,all-purpose flour,raisins,m...\n",
      "1  28583  sugar,egg yolks,corn starch,cream of tartar,ba...\n",
      "2  41580  sausage links,fennel bulb,fronds,olive oil,cub...\n",
      "3  29752  meat cuts,file powder,smoked sausage,okra,shri...\n",
      "4  35687  ground black pepper,salt,sausage casings,leeks...\n",
      "baking powder,eggs,all-purpose flour,raisins,milk,white sugar\n",
      "\n",
      "\n",
      "\n",
      "['baking powder,eggs,all-purpose flour,raisins,milk,white sugar', 'sugar,egg yolks,corn starch,cream of tartar,bananas,vanilla wafers,milk,vanilla extract,toasted pecans,egg whites,light rum']\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "          cuisine     id\n",
      "0       [british]  18009\n",
      "1   [southern_us]  28583\n",
      "2       [italian]  41580\n",
      "3  [cajun_creole]  29752\n",
      "4       [italian]  35687\n",
      "        cuisine     id\n",
      "0       british  18009\n",
      "1   southern_us  28583\n",
      "2       italian  41580\n",
      "3  cajun_creole  29752\n",
      "4       italian  35687\n"
     ]
    }
   ],
   "source": [
    "id1 = []\n",
    "ingre1 = []\n",
    "full1 = []\n",
    "for i in range(length1):\n",
    "    id1.append(test[i]['id'])\n",
    "    for j in range(len\n",
    "                   (test[i]['ingredients'])):\n",
    "        ingre1.append(test[i]['ingredients'][j])\n",
    "    li1 = [y for y in ingre1]\n",
    "    li1 = \",\".join(map(str, li1))\n",
    "    full1.append(li1)\n",
    "    ingre1 = []\n",
    "    #if(i == 5000):\n",
    "    #    break\n",
    "    if(i%5000 == 0):\n",
    "        print(i)\n",
    "        \n",
    "full2 = []\n",
    "for k in range(len(full1)):\n",
    "    full2.append(full1[k])\n",
    "\n",
    "df2 = pd.DataFrame({'id':id1,'ingredients':full1})\n",
    "print(df2.head())\n",
    "print(df2['ingredients'][0])\n",
    "ans = []\n",
    "\n",
    "print(\"\\n\\n\")\n",
    "print(full2[:2])\n",
    "print(\"\\n\\n\\n\")\n",
    "\n",
    "for k in range(len(full1)):\n",
    "    ans.append((cuis[(\\\n",
    "    lsvc.predict(count_vect.transform([full1[k]])))]))\n",
    "\n",
    "out = pd.DataFrame({'id':df2['id'], 'cuisine':ans})\n",
    "'''df2['ingredients'] = \",\".join(map(str, df2['ingredients']))\n",
    "print(df2.head())\n",
    "ans = pd.DataFrame(columns=['cuisine'])\n",
    "k = 0\n",
    "print(df2['ingredients'][0])\n",
    "print(\"\\t\\t\\tTesting\")'''\n",
    "'''for i in df2['ingredients']:\n",
    "    temp = pd.DataFrame(columns=['cuisine'])\n",
    "    temp['cuisine'] = (cuis[(\\\n",
    "    clf.predict(count_vect.transform([i])))])\n",
    "    #print(i)\n",
    "    print(temp)\n",
    "    ans = ans.append(temp)\n",
    "    k+=10\n",
    "    if(k == 1):\n",
    "        break'''\n",
    "#print(ans)\n",
    "print(out.head())\n",
    "out['cuisine'] = out['cuisine'].str[0]\n",
    "print(out.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "file = \"submission3.csv\"\n",
    "out.to_csv(file,index=False)\n",
    "#Score 0.73:"
   ]
  }
 ],
 "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": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}