w2v_svm_A.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Il notebook implementa un classificatore addestrato sulle recensioni rappresentate come vettori. I vettori vengono estratti da un Word2Vec model addestrato from scratch sul nostro dataset. I vettori delle recensioni vengono calcolati come la media ponderata vec * tf-idf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import nltk\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn import metrics\n",
    "from sklearn.svm import LinearSVC, SVC\n",
    "from sklearn.naive_bayes import MultinomialNB\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "import string\n",
    "import statistics\n",
    "from nltk import pos_tag\n",
    "from nltk.corpus import stopwords\n",
    "from nltk.corpus import wordnet\n",
    "import re\n",
    "import gensim\n",
    "from gensim.models import KeyedVectors\n",
    "import numpy as np\n",
    "from collections import Counter\n",
    "from numpy import dot\n",
    "from numpy.linalg import norm\n",
    "from gensim.models import Word2Vec\n",
    "from gensim.models.phrases import Phrases\n",
    "from nltk import sent_tokenize\n",
    "from nltk.tokenize.treebank import TreebankWordDetokenizer\n",
    "import gensim.downloader as api\n",
    "from sklearn.model_selection import GridSearchCV, StratifiedKFold\n",
    "from sklearn.metrics import accuracy_score, f1_score, classification_report\n",
    "from sklearn.metrics import accuracy_score, precision_score, recall_score, make_scorer\n",
    "from preprocessing import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "w2v_pretrained = Word2Vec.load(\"w2vPreTrained\") # carica gli embedding pre trainati sul dataset test_set.csv. ciò ci serve esclusivamente per generare recensioni sintetiche"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3150, 6)\n",
      "(3150, 6)\n",
      "(2998, 6)\n",
      "(2196, 6)\n"
     ]
    }
   ],
   "source": [
    "dataset = pd.read_csv(\"amazon_alexa.tsv\", sep = \"\\t\", encoding = \"utf-8\")\n",
    "print(dataset.shape)\n",
    "dataset.dropna(inplace = True)\n",
    "print(dataset.shape)\n",
    "dataset.drop(dataset[dataset.rating == 3].index, inplace=True)\n",
    "print(dataset.shape)\n",
    "dataset.drop_duplicates(subset = \"verified_reviews\", inplace = True)\n",
    "print(dataset.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# converte in un formato col quale lalvora imblearn\n",
    "X = np.array(dataset[\"verified_reviews\"].values).reshape(-1, 1)\n",
    "y = list(dataset[\"feedback\"].values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Resampled dataset shape Counter({1: 412, 0: 206})\n"
     ]
    }
   ],
   "source": [
    "#carichiamo il dataset e lo bilanciamo \n",
    "from imblearn.under_sampling import RandomUnderSampler\n",
    "from collections import Counter\n",
    "\n",
    "# classe negativa = 50% della classe positiva\n",
    "undersampler = RandomUnderSampler(sampling_strategy=0.5, random_state = 0)\n",
    "\n",
    "X, y = undersampler.fit_resample(X, y)\n",
    "\n",
    "\n",
    "print('Resampled dataset shape %s' % Counter(y))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_temp = []\n",
    "\n",
    "for rev in X:\n",
    "  X_temp.append(rev[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "total number of types extracted is: 1788\n"
     ]
    }
   ],
   "source": [
    "new_text, new_sent_tok = tokenize_list_of_text(X_temp, custom_stopwords, False, 2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Elimino token con freq minore di 2\n",
    "cleaned_reviews = frequency_cleaning(new_sent_tok, 2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['little', 'feature']"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Estraggo i bigrammi\n",
    "bigrams = Phrases(cleaned_reviews, scoring=\"npmi\", threshold=0.60) #estrae le collocazioni tramite PMI\n",
    "bigrams[cleaned_reviews][2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Splittiamo train e test: saranno 329-165\n",
    "X_train, X_test, Y_train, Y_test = train_test_split(bigrams[cleaned_reviews], y, test_size=0.20, random_state=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train set is Counter({1: 329, 0: 165})\n"
     ]
    }
   ],
   "source": [
    "print('Train set is %s' % Counter(Y_train))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['joke']\n",
      "-\n",
      "['laugh']\n",
      "--------------------------\n",
      "['price', 'product', 'nice', 'quality', 'nice', 'feature', 'definitely', 'reason', 'give', 'think', 'may', 'buyer', 'error', 'first', 'ignore', 'product', 'plug', 'time', 'work', 'really', 'unlike', 'not_a', 'stand', 'device', 'also', 'speaker', 'not_very', 'loud', 'buy', 'bluetooth', 'speaker', 'sure', 'lot', 'figure', 'use', 'kind', 'seem', 'like', 'device', 'plus', 'set', 'awful', 'would', 'definitely', 'money', 'buy', 'one', 'actually']\n",
      "-\n",
      "['cost', 'product', 'decent', 'quality', 'decent', 'feature', 'definitely', 'reason', 'present', 'imagine', 'may', 'buyer', 'error', 'start', 'ignore', 'product', 'plug', 'sentence', 'run', 'truly', 'different', 'not_a', 'base', 'device', 'also', 'speaker', 'not_very', 'loudly', 'purchase', 'bluetooth', 'speaker', 'sure', 'plenty', 'figure', 'utilize', 'kind', 'seem', 'wish', 'device', 'plus', 'adjust', 'terrible', 'would', 'definitely', 'money', 'purchase', 'one', 'really']\n",
      "--------------------------\n",
      "['never', 'work']\n",
      "-\n",
      "['never', 'run']\n",
      "--------------------------\n",
      "['order', 'deal', 'day', 'idle', 'day']\n",
      "-\n",
      "['order', 'bargain', 'day', 'idle', 'day']\n",
      "--------------------------\n",
      "['great', 'product', 'useless', 'overall', 'many', 'feature', 'unless', 'smart', 'home', 'obviate']\n",
      "-\n",
      "['outstanding', 'product', 'useless', 'overall', 'many', 'feature', 'unless', 'smart', 'house', 'obviate']\n",
      "--------------------------\n",
      "['use', 'idle']\n",
      "-\n",
      "['utilize', 'idle']\n",
      "--------------------------\n",
      "['much', 'difficult', 'use']\n",
      "-\n",
      "['much', 'hard', 'utilize']\n",
      "--------------------------\n",
      "['two', 'week', 'set', 'unretentive', 'work', 'dark', 'guess', 'lack', 'purchase', 'refurbished']\n",
      "-\n",
      "['two', 'week', 'adjust', 'unretentive', 'run', 'blue', 'think', 'miss', 'buy', 'refurbish']\n",
      "--------------------------\n",
      "['use', 'hear', 'news', 'concern', 'wish', 'would', 'something', 'firmware', 'yet']\n",
      "-\n",
      "['utilize', 'listen', 'news', 'worry', 'like', 'would', 'something', 'firmware', 'still']\n",
      "--------------------------\n",
      "['really', 'happy', 'original', 'thought', 'get', 'use', 'bedroom', 'really', 'disappointed', 'audio', 'quality', 'connect', 'speaker', 'via', 'bluetooth', 'audio', 'much', 'good', 'start', 'problem', 'connection', 'wifi', 'bluetooth', 'connect', 'speaker', 'via', 'wifi', 'wake', 'middle', 'night', 'horrible', 'sound', 'hop', 'take', 'thing', 'back', 'give', 'good', 'deal', 'spot', 'hope', 'good', 'device']\n",
      "-\n",
      "['truly', 'glad', 'original', 'think', 'receive', 'utilize', 'bedroom', 'truly', 'disappoint', 'sound', 'quality', 'link', 'speaker', 'via', 'bluetooth', 'sound', 'much', 'well', 'begin', 'trouble', 'connection', 'wifi', 'bluetooth', 'link', 'speaker', 'via', 'wifi', 'awake', 'eye', 'dark', 'horrible', 'audio', 'hop', 'bring', 'thing', 'back', 'present', 'well', 'bargain', 'place', 'promise', 'well', 'device']\n",
      "--------------------------\n",
      "['bad', 'purchase', 'make', 'big', 'love', 'brand', 'digital', 'music', 'video', 'even', 'buy', 'game', 'music', 'movie', 'app', 'try', 'use', 'make', 'simple', 'purchase', 'anything', 'useless']\n",
      "-\n",
      "['badly', 'buy', 'ready', 'large', 'enjoy', 'brand', 'digital', 'music', 'picture', 'still', 'purchase', 'game', 'music', 'movie', 'app', 'attempt', 'utilize', 'ready', 'simple', 'buy', 'anything', 'useless']\n",
      "--------------------------\n",
      "['sound', 'terrible', 'like', 'love', 'item', 'awful']\n",
      "-\n",
      "['audio', 'awful', 'wish', 'enjoy', 'item', 'terrible']\n",
      "--------------------------\n",
      "['product', 'two', 'software', 'make', 'completely', 'card', 'homescreen', 'call', 'thing', 'try', 'service', 'try', 'turn', 'homescreen', 'card', 'not_this', 'one', 'homescreen', 'card', 'cycle', 'annoy', 'set', 'card', 'cycle', 'instead', 'cycle', 'continuously', 'setting', 'idle', 'unit', 'set', 'cycle', 'card', 'still', 'continue', 'cycle', 'time', 'reboot', 'device', 'set', 'etc', 'etc', 'two', 'software', 'issue', 'fix', 'product', 'useless']\n",
      "-\n",
      "['product', 'two', 'software', 'ready', 'totally', 'card', 'homescreen', 'phone', 'thing', 'attempt', 'service', 'attempt', 'play', 'homescreen', 'card', 'not_this', 'one', 'homescreen', 'card', 'cycle', 'irritate', 'adjust', 'card', 'cycle', 'instead', 'cycle', 'continuously', 'position', 'idle', 'unit', 'adjust', 'cycle', 'card', 'yet', 'proceed', 'cycle', 'sentence', 'reboot', 'device', 'adjust', 'etc', 'etc', 'two', 'software', 'issue', 'repair', 'product', 'useless']\n",
      "--------------------------\n",
      "['sound', 'home', 'mini', 'sound', 'bad', 'also', 'need', 'small', 'selection', 'music', 'need', 'music', 'play', 'everything', 'also', 'get', 'two', 'need', 'family', 'music', 'play_music', 'think', 'need', 'family', 'use', 'multiple', 'device']\n",
      "-\n",
      "['audio', 'house', 'mini', 'audio', 'badly', 'also', 'want', 'little', 'option', 'music', 'want', 'music', 'turn', 'everything', 'also', 'receive', 'two', 'want', 'household', 'music', 'play_music', 'imagine', 'want', 'household', 'utilize', 'multiple', 'device']\n",
      "--------------------------\n",
      "['work', 'fine', 'realize', 'obviate', 'not_use']\n",
      "-\n",
      "['run', 'okay', 'see', 'obviate', 'not_use']\n",
      "--------------------------\n",
      "['fine', 'spot', 'exact', 'one', 'order', 'turn', 'repeat', 'alarm', 'morning', 'pretty', 'much', 'talk', 'not_a', 'thing', 'want', 'first', 'wake', 'bed', 'definitely', 'unlike', 'yes', 'swipe', 'alarm', 'face', 'dismiss', 'swipe', 'rather', 'touch', 'screen', 'half', 'asleep', 'difficult', 'set', 'snooze', 'half', 'time', 'even', 'really', 'try', 'swipe', 'dismiss', 'swipe', 'dismiss', 'completely', 'alarm', 'especially', 'annoy', 'could', 'easily', 'fix', 'make', 'touch', 'screen', 'turn', 'alarm', 'record', 'least', 'make', 'set', 'decide', 'touch', 'screen', 'alarm', 'snooze', 'turn', 'clock', 'year', 'still', 'set', 'clearly', 'may', 'big', 'product', 'ever', 'make']\n",
      "-\n",
      "['okay', 'place', 'precise', 'one', 'order', 'play', 'repeat', 'alarm', 'morning', 'fairly', 'much', 'speak', 'not_a', 'thing', 'need', 'start', 'awake', 'bed', 'definitely', 'different', 'yes', 'swipe', 'alarm', 'face', 'dismiss', 'swipe', 'rather', 'disturb', 'screen', 'half', 'asleep', 'hard', 'adjust', 'snooze', 'half', 'sentence', 'still', 'truly', 'attempt', 'swipe', 'dismiss', 'swipe', 'dismiss', 'totally', 'alarm', 'particularly', 'irritate', 'could', 'easy', 'repair', 'ready', 'disturb', 'screen', 'play', 'alarm', 'book', 'least', 'ready', 'adjust', 'settle', 'disturb', 'screen', 'alarm', 'snooze', 'play', 'clock', 'year', 'yet', 'adjust', 'clear', 'may', 'large', 'product', 'ever', 'ready']\n",
      "--------------------------\n",
      "['would', 'not_recommend', 'anyone', 'idle', 'program', 'way', 'thru', 'tell', 'unable', 'process', 'time', 'try', 'not_the', 'stream', 'fine', 'computer', 'disappointing']\n",
      "-\n",
      "['would', 'not_recommend', 'anyone', 'idle', 'program', 'way', 'thru', 'say', 'unable', 'procedure', 'sentence', 'attempt', 'not_the', 'stream', 'okay', 'computer', 'disappointing']\n",
      "--------------------------\n",
      "['product', 'dark', 'never', 'reset', 'instruction', 'actually', 'differ', 'package', 'instruction', 'try', 'product', 'defective']\n",
      "-\n",
      "['product', 'blue', 'never', 'reset', 'direction', 'really', 'differ', 'box', 'direction', 'attempt', 'product', 'faulty']\n",
      "--------------------------\n",
      "['turn']\n",
      "-\n",
      "['play']\n",
      "--------------------------\n",
      "['device', 'want', 'sync', 'phone', 'number', 'would', 'forbid', 'say', 'number', 'already', 'use', 'customer', 'service', 'could', 'basically', 'tell', 'phone', 'set', 'kind', 'pointless', 'pay']\n",
      "-\n",
      "['device', 'need', 'sync', 'call', 'amount', 'would', 'forbid', 'tell', 'amount', 'already', 'utilize', 'customer', 'service', 'could', 'essentially', 'say', 'call', 'adjust', 'kind', 'pointless', 'give']\n",
      "--------------------------\n",
      "['still', 'unable', 'connect', 'device', 'say', 'build', 'buy', 'bridge', 'connect', 'device', 'think', 'everything', 'would', 'sync', 'device', 'still', 'via', 'plus', 'look_forward', 'feature', 'able', 'play_music', 'check', 'weather', 'etc', 'fine']\n",
      "-\n",
      "['yet', 'unable', 'link', 'device', 'tell', 'build', 'purchase', 'bridge', 'link', 'device', 'imagine', 'everything', 'would', 'sync', 'device', 'yet', 'via', 'plus', 'look_forward', 'feature', 'able', 'play_music', 'check', 'weather', 'etc', 'okay']\n",
      "--------------------------\n",
      "['need', 'buy', 'get', 'work', 'many', 'device', 'want', 'money', 'back', 'thing', 'odd', 'turn', 'unless', 'buy', 'another', 'point', 'half', 'device', 'sell', 'thing', 'unless', 'want', 'spend', 'another', 'another', 'idle', 'many', 'device', 'must', 'not_this', 'good', 'yet', 'get', 'device', 'actually', 'understand', 'talk']\n",
      "-\n",
      "['want', 'purchase', 'receive', 'run', 'many', 'device', 'need', 'money', 'back', 'thing', 'odd', 'play', 'unless', 'purchase', 'another', 'point', 'half', 'device', 'trade', 'thing', 'unless', 'need', 'spend', 'another', 'another', 'idle', 'many', 'device', 'must', 'not_this', 'well', 'still', 'receive', 'device', 'really', 'interpret', 'speak']\n",
      "--------------------------\n",
      "['replace', 'one', 'move', 'lack', 'range', 'refurbish', 'unit', 'would', 'able', 'find', 'could', 'disconnect', 'get', 'hear', 'trouble', 'connect', 'check', 'app', 'unplug', 'several', 'time', 'not_a']\n",
      "-\n",
      "['substitute', 'one', 'movement', 'miss', 'range', 'refurbish', 'unit', 'would', 'able', 'discover', 'could', 'unplug', 'receive', 'listen', 'problem', 'link', 'check', 'app', 'disconnect', 'various', 'sentence', 'not_a']\n",
      "--------------------------\n",
      "['purchase', 'use', 'idle', 'application']\n",
      "-\n",
      "['buy', 'utilize', 'idle', 'application']\n",
      "--------------------------\n",
      "['much', 'price', 'speaker']\n",
      "-\n",
      "['much', 'cost', 'speaker']\n",
      "--------------------------\n",
      "['suppose', 'come', 'free', 'hue', 'day', 'dark', 'bulb']\n",
      "-\n",
      "['think', 'arrive', 'free', 'hue', 'day', 'blue', 'lightbulb']\n",
      "--------------------------\n",
      "['idle', 'say', 'disconnect', 'network']\n",
      "-\n",
      "['idle', 'tell', 'unplug', 'mesh']\n",
      "--------------------------\n",
      "['buy', 'refurbish', 'stop', 'function', 'quite', 'often', 'sleep', 'every', 'time', 'randomly', 'turn', 'nothing', 'sell', 'refurbished']\n",
      "-\n",
      "['purchase', 'refurbish', 'quit', 'purpose', 'quite', 'frequently', 'sleep', 'every', 'sentence', 'randomly', 'play', 'nothing', 'trade', 'refurbish']\n",
      "--------------------------\n",
      "['product', 'complete', 'waste', 'money', 'ask', 'question', 'either', 'ignore', 'answer', 'give', 'wrong', 'answer', 'find', 'information']\n",
      "-\n",
      "['product', 'finish', 'waste', 'money', 'demand', 'inquiry', 'either', 'ignore', 'response', 'present', 'wrong', 'response', 'discover', 'info']\n",
      "--------------------------\n",
      "['thing', 'work', 'party', 'apps', 'stuff', 'think', 'could', 'box', 'send_back', 'waste', 'money']\n",
      "-\n",
      "['thing', 'run', 'party', 'apps', 'stuff', 'imagine', 'could', 'package', 'send_back', 'waste', 'money']\n",
      "--------------------------\n",
      "['second', 'one', 'refurbished', 'work', 'least', 'home']\n",
      "-\n",
      "['minute', 'one', 'refurbish', 'run', 'least', 'house']\n",
      "--------------------------\n",
      "['item', 'never', 'work', 'box', 'spent', 'several', 'day', 'try', 'get', 'work', 'run', 'fix', 'thing', 'never', 'order', 'another', 'refurbish', 'device']\n",
      "-\n",
      "['item', 'never', 'run', 'package', 'spend', 'various', 'day', 'attempt', 'receive', 'run', 'work', 'repair', 'thing', 'never', 'order', 'another', 'refurbish', 'device']\n",
      "--------------------------\n",
      "['even_though', 'say', 'would', 'spend', 'multiple', 'hour', 'make', 'phone', 'call', 'also', 'miss', 'free', 'bulb']\n",
      "-\n",
      "['even_though', 'tell', 'would', 'spend', 'multiple', 'minute', 'ready', 'call', 'phone', 'also', 'lack', 'free', 'lightbulb']\n",
      "--------------------------\n",
      "['ignore', 'answer', 'many', 'question', 'ask', 'example', 'ask', 'many', 'say', 'money']\n",
      "-\n",
      "['ignore', 'response', 'many', 'inquiry', 'demand', 'example', 'demand', 'many', 'tell', 'money']\n",
      "--------------------------\n",
      "['work', 'well', 'month', 'stop', 'connect', 'internet', 'wifi', 'trouble', 'work', 'customer', 'service', 'could', 'far', 'tell', 'month', 'disappointing', 'replace']\n",
      "-\n",
      "['run', 'good', 'month', 'quit', 'link', 'internet', 'wifi', 'problem', 'run', 'customer', 'service', 'could', 'far', 'say', 'month', 'disappointing', 'substitute']\n",
      "--------------------------\n",
      "['receive', 'wrong', 'product', 'install', 'excitement', 'thank']\n",
      "-\n",
      "['get', 'wrong', 'product', 'instal', 'excitement', 'thank']\n",
      "--------------------------\n",
      "['like', 'day', 'year', 'unretentive', 'work', 'would', 'inability', 'past', 'buy', 'second', 'one', 'sell', 'third', 'one', 'first', 'longer']\n",
      "-\n",
      "['wish', 'day', 'year', 'unretentive', 'run', 'would', 'inability', 'past', 'purchase', 'minute', 'one', 'trade', 'third', 'one', 'start', 'longer']\n",
      "--------------------------\n",
      "['waste', 'money', 'not_really', 'get', 'anything', 'video_chat', 'device', 'would', 'picture', 'put', 'mode', 'first', 'photo', 'even', 'look', 'research', 'lose', 'also', 'problem', 'device', 'lack', 'problem', 'device', 'price', 'difference', 'worthlessness']\n",
      "-\n",
      "['waste', 'money', 'not_really', 'receive', 'anything', 'video_chat', 'device', 'would', 'photo', 'place', 'mood', 'start', 'picture', 'still', 'see', 'research', 'lose', 'also', 'trouble', 'device', 'miss', 'trouble', 'device', 'cost', 'difference', 'worthlessness']\n",
      "--------------------------\n",
      "['disappointed', 'request']\n",
      "-\n",
      "['disappoint', 'request']\n",
      "--------------------------\n",
      "['item', 'return', 'repair', 'item', 'back', 'repair', 'part', 'miss', 'inability', 'cord', 'include', 'please']\n",
      "-\n",
      "['item', 'refund', 'fix', 'item', 'back', 'fix', 'portion', 'lack', 'inability', 'cord', 'include', 'please']\n",
      "--------------------------\n",
      "['send_back', 'turn', 'expect', 'refurbished', 'item']\n",
      "-\n",
      "['send_back', 'play', 'anticipate', 'refurbish', 'item']\n",
      "--------------------------\n",
      "['bad', 'device', 'ignore', 'refurbish', 'activate', 'talk', 'activate', 'start', 'talk', 'talk', 'super']\n",
      "-\n",
      "['badly', 'device', 'ignore', 'refurbish', 'trigger', 'speak', 'trigger', 'begin', 'speak', 'speak', 'extremely']\n",
      "--------------------------\n",
      "['get', 'month', 'idea', 'help', 'control', 'smart', 'bulb', 'useful', 'answer_question', 'get', 'daily', 'weather', 'update', 'far', 'useful', 'half', 'time', 'wake', 'device', 'ask', 'question', 'command', 'say', 'turn', 'basically', 'ignore', 'often', 'unable', 'answer_question', 'work', 'rather', 'make_life', 'easy', 'would', 'return', 'past', 'return']\n",
      "-\n",
      "['receive', 'month', 'idea', 'assistance', 'operate', 'smart', 'lightbulb', 'useful', 'answer_question', 'receive', 'everyday', 'weather', 'update', 'far', 'useful', 'half', 'sentence', 'awake', 'device', 'demand', 'inquiry', 'control', 'tell', 'play', 'essentially', 'ignore', 'frequently', 'unable', 'answer_question', 'run', 'rather', 'make_life', 'easily', 'would', 'refund', 'past', 'refund']\n",
      "--------------------------\n",
      "['slow']\n",
      "-\n",
      "['sluggish']\n",
      "--------------------------\n",
      "['differ', 'excite', 'try', 'new', 'unfortunately', 'even_though', 'foot', 'away', 'router', 'get', 'think', 'little', 'device', 'suppose', 'take', 'issue']\n",
      "-\n",
      "['differ', 'excite', 'attempt', 'new', 'unfortunately', 'even_though', 'foot', 'away', 'router', 'receive', 'imagine', 'small', 'device', 'think', 'bring', 'issue']\n",
      "--------------------------\n",
      "['get', 'excite', 'device', 'good', 'sound', 'control', 'hue', 'not_realize', 'also', 'communicate', 'not_use', 'hue', 'bulb', 'full', 'potential', 'device', 'turn', 'light', 'set', 'number', 'color', 'also', 'bulb', 'new', 'security', 'concern', 'sell', 'want', 'full', 'new', 'device', 'end', 'need', 'update', 'make', 'device', 'pointless', 'still', 'love', 'complaint', 'functionality', 'sound_quality', 'device', 'disappointed', 'functionality']\n",
      "-\n",
      "['receive', 'excite', 'device', 'well', 'audio', 'operate', 'hue', 'not_realize', 'also', 'pass', 'not_use', 'hue', 'lightbulb', 'wide', 'potential', 'device', 'play', 'lighting', 'adjust', 'amount', 'color', 'also', 'lightbulb', 'new', 'security', 'worry', 'trade', 'need', 'wide', 'new', 'device', 'end', 'want', 'update', 'ready', 'device', 'pointless', 'yet', 'enjoy', 'complaint', 'functionality', 'sound_quality', 'device', 'disappoint', 'functionality']\n",
      "--------------------------\n",
      "['get', 'try', 'screen']\n",
      "-\n",
      "['receive', 'attempt', 'screen']\n",
      "--------------------------\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['purchase', 'plus', 'build', 'come', 'free', 'hue', 'bulb', 'get', 'everything', 'set', 'realize', 'plus', 'forbid', 'full', 'functionality', 'bulb', 'turn', 'light', 'stay', 'color', 'still', 'hue', 'bridge', 'full', 'function', 'feature', 'bulb', 'wish', 'buy', 'bulb', 'bridge', 'use', 'already', 'since', 'build', 'useless', 'buy', 'hue', 'bridge', 'get', 'full', 'light']\n",
      "-\n",
      "['buy', 'plus', 'build', 'arrive', 'free', 'hue', 'lightbulb', 'receive', 'everything', 'adjust', 'see', 'plus', 'forbid', 'wide', 'functionality', 'lightbulb', 'play', 'lighting', 'stop', 'color', 'yet', 'hue', 'bridge', 'wide', 'purpose', 'feature', 'lightbulb', 'like', 'purchase', 'lightbulb', 'bridge', 'utilize', 'already', 'since', 'build', 'useless', 'purchase', 'hue', 'bridge', 'receive', 'wide', 'lighting']\n",
      "--------------------------\n",
      "['terrible', 'stop', 'work', 'one', 'day']\n",
      "-\n",
      "['awful', 'quit', 'run', 'one', 'day']\n",
      "--------------------------\n",
      "['old', 'keep', 'crash', 'use', 'buy', 'one', 'think', 'old', 'one', 'defective', 'one', 'thing', 'back', 'start', 'leave', 'program']\n",
      "-\n",
      "['previous', 'prevent', 'crash', 'utilize', 'purchase', 'one', 'imagine', 'previous', 'one', 'faulty', 'one', 'thing', 'back', 'begin', 'forget', 'program']\n",
      "--------------------------\n",
      "['unbend', 'thing', 'try', 'screen', 'scroll', 'big', 'device', 'send_back', 'really', 'want', 'get', 'enough', 'already']\n",
      "-\n",
      "['unbend', 'thing', 'attempt', 'screen', 'scroll', 'large', 'device', 'send_back', 'truly', 'need', 'receive', 'enough', 'already']\n",
      "--------------------------\n",
      "['never', 'respond', 'speak', 'word', 'think', 'would', 'connect', 'use', 'like', 'intercom', 'differ', 'able']\n",
      "-\n",
      "['never', 'answer', 'talk', 'word', 'imagine', 'would', 'link', 'utilize', 'wish', 'intercom', 'differ', 'able']\n",
      "--------------------------\n",
      "['poor', 'quality', 'give', 'away']\n",
      "-\n",
      "['poor', 'quality', 'present', 'away']\n",
      "--------------------------\n",
      "['return', 'piece', 'garbage', 'soon', 'sure', 'small', 'customer', 'rarely', 'disappointed', 'purchase', 'home', 'automation', 'buy', 'plus', 'build', 'along', 'zigbee', 'light', 'switch', 'voice', 'control', 'even', 'work', 'often', 'nothing', 'could', 'differ', 'disappointed', 'product', 'zigbee', 'value', 'money', 'expect', 'product', 'need', 'look']\n",
      "-\n",
      "['refund', 'piece', 'garbage', 'shortly', 'sure', 'little', 'customer', 'seldom', 'disappoint', 'buy', 'house', 'automation', 'purchase', 'plus', 'build', 'along', 'zigbee', 'lighting', 'change', 'voice', 'operate', 'still', 'run', 'frequently', 'nothing', 'could', 'differ', 'disappoint', 'product', 'zigbee', 'value', 'money', 'anticipate', 'product', 'want', 'see']\n",
      "--------------------------\n",
      "['sound_quality', 'not_that', 'good', 'especially', 'price', 'pay', 'would', 'recommend', 'get', 'buy', 'good', 'speaker', 'instead', 'maybe', 'sonos', 'use']\n",
      "-\n",
      "['sound_quality', 'not_that', 'well', 'particularly', 'cost', 'give', 'would', 'recommend', 'receive', 'purchase', 'well', 'speaker', 'instead', 'perhaps', 'sonos', 'utilize']\n",
      "--------------------------\n",
      "['think', 'get', 'long', 'time', 'excite', 'hue', 'bulb', 'excite', 'end', 'spend', 'hour', 'attempt', 'add', 'bulb', 'plus', 'everything', 'try', 'end', 'try', 'set', 'watch', 'video', 'read', 'set', 'thing', 'expect', 'different', 'next', 'day', 'get', 'tech', 'support', 'hop', 'something', 'would', 'make', 'thing', 'work', 'also', 'problem', 'differ', 'error', 'well', 'tech', 'support', 'phone', 'still', 'idle', 'product', 'test', 'far', 'experience', 'smart', 'home', 'old', 'school', 'reason', 'give', 'two', 'send', 'different', 'plus', 'light_bulb', 'problem', 'not_the', 'bulb', 'item', 'purchase', 'another', 'bulb', 'still', 'could', 'device', 'next', 'attempt', 'defective', 'support', 'thought', 'maybe', 'explore', 'device', 'end', 'look', 'look', 'like', 'disbelieve', 'lose', 'video', 'show', 'problem', 'everything', 'look', 'simple', 'not_the', 'experience', 'review', 'spend', 'enough', 'time', 'already']\n",
      "-\n",
      "['imagine', 'receive', 'long', 'sentence', 'excite', 'hue', 'lightbulb', 'excite', 'end', 'spend', 'minute', 'try', 'bring', 'lightbulb', 'plus', 'everything', 'attempt', 'end', 'attempt', 'adjust', 'see', 'picture', 'understand', 'adjust', 'thing', 'anticipate', 'unlike', 'future', 'day', 'receive', 'tech', 'support', 'hop', 'something', 'would', 'ready', 'thing', 'run', 'also', 'trouble', 'differ', 'error', 'good', 'tech', 'support', 'call', 'yet', 'idle', 'product', 'test', 'far', 'experience', 'smart', 'house', 'previous', 'train', 'reason', 'present', 'two', 'ship', 'unlike', 'plus', 'light_bulb', 'trouble', 'not_the', 'lightbulb', 'item', 'buy', 'another', 'lightbulb', 'yet', 'could', 'device', 'future', 'try', 'faulty', 'support', 'think', 'perhaps', 'explore', 'device', 'end', 'see', 'see', 'wish', 'disbelieve', 'lose', 'picture', 'picture', 'trouble', 'everything', 'see', 'simple', 'not_the', 'experience', 'review', 'spend', 'enough', 'sentence', 'already']\n",
      "--------------------------\n",
      "['horrible', 'volume', 'phone', 'loud', 'device', 'one', 'not_the', 'case']\n",
      "-\n",
      "['horrible', 'loudness', 'call', 'loudly', 'device', 'one', 'not_the', 'case']\n",
      "--------------------------\n",
      "['come']\n",
      "-\n",
      "['arrive']\n",
      "--------------------------\n",
      "['sound', 'terrible', 'want', 'good', 'music', 'get']\n",
      "-\n",
      "['audio', 'awful', 'need', 'well', 'music', 'receive']\n",
      "--------------------------\n",
      "['speaker', 'pretty', 'terrible', 'home', 'good', 'product']\n",
      "-\n",
      "['speaker', 'fairly', 'awful', 'house', 'well', 'product']\n",
      "--------------------------\n",
      "['wish', 'battery', 'well', 'plug', 'item']\n",
      "-\n",
      "['like', 'battery', 'good', 'plug', 'item']\n",
      "--------------------------\n",
      "['idle', 'time']\n",
      "-\n",
      "['idle', 'sentence']\n",
      "--------------------------\n",
      "['smart', 'enough', 'make', 'work', 'send_back', 'cant', 'get', 'help', 'work', 'fine', 'want', 'big', 'speaker']\n",
      "-\n",
      "['smart', 'enough', 'ready', 'run', 'send_back', 'bank', 'receive', 'assistance', 'run', 'okay', 'need', 'large', 'speaker']\n",
      "--------------------------\n",
      "['ignore', 'customer', 'use', 'device', 'tool', 'screen', 'unbend', 'thing', 'try', 'people', 'turn', 'change', 'option', 'size', 'slightly', 'improve', 'audio', 'speaker', 'would', 'give', 'device', 'right', 'return', 'device', 'wife', 'say']\n",
      "-\n",
      "['ignore', 'customer', 'utilize', 'device', 'tool', 'screen', 'unbend', 'thing', 'attempt', 'people', 'play', 'switch', 'choice', 'size', 'slightly', 'improve', 'sound', 'speaker', 'would', 'present', 'device', 'right', 'refund', 'device', 'wife', 'tell']\n",
      "--------------------------\n",
      "['think', 'really', 'unmake', 'much', 'play_music', 'answer', 'stupid', 'question', 'finally', 'able', 'hook', 'thermostat', 'see', 'price', 'drop', 'like']\n",
      "-\n",
      "['imagine', 'truly', 'unmake', 'much', 'play_music', 'response', 'stupid', 'inquiry', 'eventually', 'able', 'addict', 'thermostat', 'visit', 'cost', 'drop', 'wish']\n",
      "--------------------------\n",
      "['buy', 'son', 'idle', 'return']\n",
      "-\n",
      "['purchase', 'son', 'idle', 'refund']\n",
      "--------------------------\n",
      "['first', 'one', 'work', 'get', 'next', 'one', 'voice', 'time', 'way', 'get', 'work', 'app', 'uncertain', 'way']\n",
      "-\n",
      "['start', 'one', 'run', 'receive', 'future', 'one', 'voice', 'sentence', 'way', 'receive', 'run', 'app', 'uncertain', 'way']\n",
      "--------------------------\n",
      "['make', 'receive', 'call', 'device', 'tell', 'need', 'ignore']\n",
      "-\n",
      "['ready', 'get', 'phone', 'device', 'say', 'want', 'ignore']\n",
      "--------------------------\n",
      "['quite', 'disappointed', 'product', 'clearly', 'half', 'time', 'ask', 'two', 'second', 'start', 'whole', 'thing', 'stop', 'couple', 'minute', 'ask', 'play_radio', 'also', 'many', 'time', 'get', 'ask', 'switch', 'light', 'item', 'also', 'name', 'clearly', 'still']\n",
      "-\n",
      "['quite', 'disappoint', 'product', 'clear', 'half', 'sentence', 'demand', 'two', 'minute', 'begin', 'totally', 'thing', 'quit', 'couple', 'second', 'demand', 'play_radio', 'also', 'many', 'sentence', 'receive', 'demand', 'change', 'lighting', 'item', 'also', 'call', 'clear', 'yet']\n",
      "--------------------------\n",
      "['unhappy', 'spot', 'reset', 'time', 'seem', 'slow', 'mode', 'top', 'hear', 'even', 'talk', 'support', 'still', 'lack', 'choose']\n",
      "-\n",
      "['unhappy', 'place', 'reset', 'sentence', 'seem', 'sluggish', 'mood', 'cover', 'listen', 'still', 'speak', 'support', 'yet', 'miss', 'select']\n",
      "--------------------------\n",
      "['load', 'keep', 'reboot', 'middle', 'show']\n",
      "-\n",
      "['load', 'prevent', 'reboot', 'eye', 'picture']\n",
      "--------------------------\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['device', 'show', 'seem', 'huge', 'potential', 'really', 'want', 'like', 'soon', 'find', 'quite', 'bit', 'disappointing', 'not_the', 'end', 'world', 'use', 'like', 'use', 'little', 'month', 'screen', 'start', 'screen', 'quick', 'look', 'make', 'understand', 'pointless', 'ask', 'replacement', 'people', 'device', 'replace', 'start', 'answer', 'quite', 'time', 'long', 'story', 'short', 'move', 'use', 'expensive', 'voice', 'activate', 'light', 'switch', 'hue', 'light', 'like']\n",
      "-\n",
      "['device', 'picture', 'seem', 'vast', 'potential', 'truly', 'need', 'wish', 'shortly', 'discover', 'quite', 'bit', 'disappointing', 'not_the', 'end', 'world', 'utilize', 'wish', 'utilize', 'small', 'month', 'screen', 'begin', 'screen', 'immediate', 'see', 'ready', 'interpret', 'pointless', 'demand', 'replacement', 'people', 'device', 'substitute', 'begin', 'response', 'quite', 'sentence', 'long', 'story', 'dead', 'movement', 'utilize', 'expensive', 'voice', 'trigger', 'lighting', 'change', 'hue', 'lighting', 'wish']\n",
      "--------------------------\n",
      "['not_it', 'end', 'help', 'replace']\n",
      "-\n",
      "['not_it', 'end', 'assistance', 'substitute']\n",
      "--------------------------\n",
      "['sound', 'terrible', 'play', 'must', 'upgrade', 'music', 'want', 'play', 'one', 'device', 'time']\n",
      "-\n",
      "['audio', 'awful', 'turn', 'must', 'upgrade', 'music', 'need', 'turn', 'one', 'device', 'sentence']\n",
      "--------------------------\n",
      "['need', 'able', 'connect', 'party', 'apps']\n",
      "-\n",
      "['want', 'able', 'link', 'party', 'apps']\n",
      "--------------------------\n",
      "['work', 'fine', 'get', 'thing', 'try', 'not_a', 'year', 'old', 'need', 'tell', 'thing', 'find', 'not_way', 'turn', 'except', 'turn', 'screen', 'lose', 'fix', 'back']\n",
      "-\n",
      "['run', 'okay', 'receive', 'thing', 'attempt', 'not_a', 'year', 'previous', 'want', 'say', 'thing', 'discover', 'not_way', 'play', 'except', 'play', 'screen', 'lose', 'repair', 'back']\n",
      "--------------------------\n"
     ]
    }
   ],
   "source": [
    "# genera recensioni sintetiche a partire dal train e usato solo nell'addestramento.\n",
    "#prendo solo il train, non prendo ne recensioni artificiali, ne le prendo dal train, che sarebbe troppo facile.\n",
    "#faccio zip di xtrain,ytrain e guardo se è negativa me la appendo alla lista delle negative\n",
    "negative_reviews = []\n",
    "for rev, label in zip(X_train, Y_train):\n",
    "    if label == 0:\n",
    "        negative_reviews.append(rev)\n",
    "\n",
    "#\n",
    "generated_reviews = generate_samples(negative_reviews, int(len(negative_reviews)/2), w2v_pretrained)\n",
    "\n",
    "# run this only one time for extend. It shows some examples real vs artificial:\n",
    "X_train.extend(generated_reviews)\n",
    "Y_train.extend([0 for x in generated_reviews])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train set is Counter({1: 329, 0: 241})\n"
     ]
    }
   ],
   "source": [
    "print('Train set is %s' % Counter(Y_train))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "bass 0.43857828532933923\n",
      "get 0.22024499174682433\n",
      "great 0.2247690722731819\n",
      "idle 0.29911950316887775\n",
      "miss 0.36554410475419424\n",
      "recognize 0.4556341840374841\n",
      "sound 0.26570121995060786\n",
      "still 0.2855958668417301\n",
      "thing 0.24613904269023423\n",
      "time 0.25529600517955614\n"
     ]
    }
   ],
   "source": [
    "#tfidf weighting sui token\n",
    "# lambda: i:i evita tokenizzazione\n",
    "\n",
    "tfidf = TfidfVectorizer(tokenizer=lambda i:i, lowercase=False, min_df = 0)\n",
    "tfidf_model = tfidf.fit_transform(X_train) #\n",
    "X_test_tf = tfidf.transform(X_test) #\n",
    "review = 1\n",
    "for score, feature in zip(tfidf_model.toarray()[review], tfidf.get_feature_names_out()):\n",
    "    if score > 0.0:\n",
    "        print(feature, score)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['purchase', 'new', 'like', 'look', 'box', 'original',\n",
       "       'even_though', 'perfectly', 'work', 'far', 'week', 'two',\n",
       "       'refurbish', 'happy'], dtype='<U15')"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tfidf.inverse_transform(tfidf_model)[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.2030027  0.2146646  0.16414036 0.22487128 0.28412446 0.26694879\n",
      " 0.27777188 0.29964046 0.14736372 0.21283613 0.27208931 0.2374712\n",
      " 0.2295642  0.51587761]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array(['purchase', 'new', 'like', 'look', 'box', 'original',\n",
       "       'even_though', 'perfectly', 'work', 'far', 'week', 'two',\n",
       "       'refurbish', 'happy'], dtype='<U15')"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(tfidf_model[0,:].data)\n",
    "tfidf.inverse_transform(tfidf_model)[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Word2Vec: \n",
    "rappresentazione vettoriale delle parole in modo che parole semanticamente simili siano rappresentate da vettori simili.\n",
    "L'algoritmo word2vec utilizza una rete neurale artificiale per imparare le relazioni tra le parole e crea un modello che prevede la probabilità che una determinata parola appaia in una data finestra di contesto. I vettori di parole creati da word2vec sono noti come \"word embeddings\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "#add only train\n",
    "#skipgram: prev c, dato p, cbow\n",
    "w2v_model = Word2Vec(X_train, vector_size=100, window = 10, min_count = 0, sg=1, hs = 1, epochs= 100, seed = 5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('great', 0.4427292048931122),\n",
       " ('use', 0.4180808961391449),\n",
       " ('cook', 0.4167807102203369),\n",
       " ('trailer', 0.38419869542121887),\n",
       " ('list', 0.3723585307598114),\n",
       " ('movie', 0.37192484736442566),\n",
       " ('find', 0.3699047267436981),\n",
       " ('want', 0.36399388313293457),\n",
       " ('favorite', 0.3593404293060303),\n",
       " ('live', 0.35832124948501587)]"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "w2v_model.wv.most_similar(\"love\", topn = 10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('unplug', -0.06076349318027496),\n",
       " ('money', -0.06643775850534439),\n",
       " ('listen', -0.06783126294612885),\n",
       " ('phone', -0.07216164469718933),\n",
       " ('zigbee', -0.07591379433870316),\n",
       " ('useful', -0.07745841145515442),\n",
       " ('half', -0.07897010445594788),\n",
       " ('return', -0.08645745366811752),\n",
       " ('ignore', -0.11667902767658234),\n",
       " ('also', -0.13518528640270233)]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "w2v_model.wv.most_similar(\"love\", topn = 2000)[-10:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "def review_vectors(tokens, size = 100, weights = []):\n",
    "\n",
    "    \"\"\"Genera un vettore per ogni recensione: questo vettore\n",
    "    è calcolato come la media ponderata (t * w: token vettore myfts * peso tfidf) dei vettori dei token nella recensione.\n",
    "    Il vettore risultante è normalizzato alla fine.\"\"\"\n",
    "\n",
    "    vec = np.zeros(size).reshape((1, size))\n",
    "    count = 0\n",
    "\n",
    "    import statistics\n",
    "\n",
    "    # if len(tokens) != len(weights):\n",
    "    #     print(\"nope\")\n",
    "\n",
    "    for word, weight in zip(tokens, weights):\n",
    "        try:\n",
    "            vec += w2v_model.wv[word] * weight #somm pond\n",
    "            count +=1 #sommiamo a vec\n",
    "        except KeyError:\n",
    "            print(\"non trovo\", word)\n",
    "            continue\n",
    "\n",
    "    if count!= 0:\n",
    "        vec = vec / norm(vec) \n",
    "        \n",
    "    return vec"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>90</th>\n",
       "      <th>91</th>\n",
       "      <th>92</th>\n",
       "      <th>93</th>\n",
       "      <th>94</th>\n",
       "      <th>95</th>\n",
       "      <th>96</th>\n",
       "      <th>97</th>\n",
       "      <th>98</th>\n",
       "      <th>99</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-0.082683</td>\n",
       "      <td>-0.151384</td>\n",
       "      <td>-0.074864</td>\n",
       "      <td>0.037651</td>\n",
       "      <td>-0.066716</td>\n",
       "      <td>0.020628</td>\n",
       "      <td>-0.011856</td>\n",
       "      <td>-0.013399</td>\n",
       "      <td>0.176482</td>\n",
       "      <td>-0.062232</td>\n",
       "      <td>...</td>\n",
       "      <td>0.260981</td>\n",
       "      <td>-0.014668</td>\n",
       "      <td>-0.199746</td>\n",
       "      <td>0.059677</td>\n",
       "      <td>-0.167955</td>\n",
       "      <td>0.148126</td>\n",
       "      <td>0.021636</td>\n",
       "      <td>-0.011042</td>\n",
       "      <td>0.135225</td>\n",
       "      <td>-0.107625</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.010347</td>\n",
       "      <td>0.124159</td>\n",
       "      <td>-0.066431</td>\n",
       "      <td>-0.006549</td>\n",
       "      <td>0.084815</td>\n",
       "      <td>0.066802</td>\n",
       "      <td>-0.015166</td>\n",
       "      <td>-0.080459</td>\n",
       "      <td>0.349125</td>\n",
       "      <td>0.013353</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.020182</td>\n",
       "      <td>-0.003024</td>\n",
       "      <td>0.030451</td>\n",
       "      <td>0.011098</td>\n",
       "      <td>-0.092784</td>\n",
       "      <td>-0.092140</td>\n",
       "      <td>-0.167290</td>\n",
       "      <td>0.066240</td>\n",
       "      <td>0.080714</td>\n",
       "      <td>-0.131729</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.081133</td>\n",
       "      <td>0.065063</td>\n",
       "      <td>0.020470</td>\n",
       "      <td>0.089003</td>\n",
       "      <td>0.026025</td>\n",
       "      <td>0.060736</td>\n",
       "      <td>-0.083863</td>\n",
       "      <td>-0.036441</td>\n",
       "      <td>0.224956</td>\n",
       "      <td>-0.214574</td>\n",
       "      <td>...</td>\n",
       "      <td>0.027686</td>\n",
       "      <td>-0.022007</td>\n",
       "      <td>-0.130763</td>\n",
       "      <td>0.073864</td>\n",
       "      <td>-0.040541</td>\n",
       "      <td>0.041402</td>\n",
       "      <td>-0.262502</td>\n",
       "      <td>-0.102487</td>\n",
       "      <td>-0.001430</td>\n",
       "      <td>-0.006950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.020947</td>\n",
       "      <td>-0.121957</td>\n",
       "      <td>0.086852</td>\n",
       "      <td>0.145406</td>\n",
       "      <td>-0.006548</td>\n",
       "      <td>0.070339</td>\n",
       "      <td>0.088396</td>\n",
       "      <td>-0.196170</td>\n",
       "      <td>0.179448</td>\n",
       "      <td>-0.215863</td>\n",
       "      <td>...</td>\n",
       "      <td>0.096574</td>\n",
       "      <td>-0.030186</td>\n",
       "      <td>-0.172965</td>\n",
       "      <td>-0.064311</td>\n",
       "      <td>0.070436</td>\n",
       "      <td>0.087530</td>\n",
       "      <td>0.010788</td>\n",
       "      <td>-0.023226</td>\n",
       "      <td>-0.024420</td>\n",
       "      <td>-0.023326</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.007948</td>\n",
       "      <td>-0.134698</td>\n",
       "      <td>-0.004376</td>\n",
       "      <td>0.008096</td>\n",
       "      <td>0.054369</td>\n",
       "      <td>-0.095736</td>\n",
       "      <td>-0.063928</td>\n",
       "      <td>-0.149390</td>\n",
       "      <td>0.069077</td>\n",
       "      <td>-0.160213</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.074355</td>\n",
       "      <td>-0.044783</td>\n",
       "      <td>-0.103684</td>\n",
       "      <td>0.085555</td>\n",
       "      <td>-0.113270</td>\n",
       "      <td>0.023193</td>\n",
       "      <td>-0.009795</td>\n",
       "      <td>-0.042300</td>\n",
       "      <td>0.032336</td>\n",
       "      <td>-0.023733</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>565</th>\n",
       "      <td>0.092886</td>\n",
       "      <td>-0.049526</td>\n",
       "      <td>0.126560</td>\n",
       "      <td>0.222132</td>\n",
       "      <td>0.004570</td>\n",
       "      <td>0.009963</td>\n",
       "      <td>0.064991</td>\n",
       "      <td>-0.123555</td>\n",
       "      <td>0.256326</td>\n",
       "      <td>-0.092037</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.023978</td>\n",
       "      <td>-0.111858</td>\n",
       "      <td>0.000663</td>\n",
       "      <td>0.030032</td>\n",
       "      <td>0.011269</td>\n",
       "      <td>0.137639</td>\n",
       "      <td>0.026717</td>\n",
       "      <td>0.129009</td>\n",
       "      <td>0.083214</td>\n",
       "      <td>0.049911</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>566</th>\n",
       "      <td>-0.033430</td>\n",
       "      <td>-0.122807</td>\n",
       "      <td>0.129224</td>\n",
       "      <td>0.215387</td>\n",
       "      <td>-0.024991</td>\n",
       "      <td>0.105416</td>\n",
       "      <td>0.078930</td>\n",
       "      <td>-0.059992</td>\n",
       "      <td>0.177580</td>\n",
       "      <td>-0.024562</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.019837</td>\n",
       "      <td>-0.163467</td>\n",
       "      <td>-0.095012</td>\n",
       "      <td>-0.091858</td>\n",
       "      <td>-0.104963</td>\n",
       "      <td>-0.028798</td>\n",
       "      <td>-0.044007</td>\n",
       "      <td>0.096463</td>\n",
       "      <td>0.089490</td>\n",
       "      <td>0.086948</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>567</th>\n",
       "      <td>-0.062368</td>\n",
       "      <td>0.007354</td>\n",
       "      <td>-0.040946</td>\n",
       "      <td>0.074390</td>\n",
       "      <td>0.186154</td>\n",
       "      <td>-0.036132</td>\n",
       "      <td>0.062150</td>\n",
       "      <td>0.035436</td>\n",
       "      <td>0.211906</td>\n",
       "      <td>-0.113501</td>\n",
       "      <td>...</td>\n",
       "      <td>0.062830</td>\n",
       "      <td>-0.012916</td>\n",
       "      <td>0.028467</td>\n",
       "      <td>-0.000391</td>\n",
       "      <td>-0.112686</td>\n",
       "      <td>0.147133</td>\n",
       "      <td>-0.004725</td>\n",
       "      <td>-0.102680</td>\n",
       "      <td>0.070786</td>\n",
       "      <td>-0.056791</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>568</th>\n",
       "      <td>-0.167305</td>\n",
       "      <td>0.016464</td>\n",
       "      <td>-0.089212</td>\n",
       "      <td>0.076259</td>\n",
       "      <td>-0.068883</td>\n",
       "      <td>0.175772</td>\n",
       "      <td>-0.087274</td>\n",
       "      <td>-0.135189</td>\n",
       "      <td>0.215763</td>\n",
       "      <td>0.032052</td>\n",
       "      <td>...</td>\n",
       "      <td>0.051127</td>\n",
       "      <td>-0.065445</td>\n",
       "      <td>0.104959</td>\n",
       "      <td>0.203136</td>\n",
       "      <td>-0.040568</td>\n",
       "      <td>0.077459</td>\n",
       "      <td>0.030946</td>\n",
       "      <td>-0.037723</td>\n",
       "      <td>0.022651</td>\n",
       "      <td>-0.017791</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>569</th>\n",
       "      <td>0.020393</td>\n",
       "      <td>-0.104050</td>\n",
       "      <td>0.084831</td>\n",
       "      <td>-0.064445</td>\n",
       "      <td>-0.149178</td>\n",
       "      <td>0.045278</td>\n",
       "      <td>0.170535</td>\n",
       "      <td>-0.069695</td>\n",
       "      <td>0.126067</td>\n",
       "      <td>0.032132</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.004386</td>\n",
       "      <td>-0.062179</td>\n",
       "      <td>-0.004828</td>\n",
       "      <td>0.000677</td>\n",
       "      <td>-0.062164</td>\n",
       "      <td>0.117224</td>\n",
       "      <td>-0.069817</td>\n",
       "      <td>0.033966</td>\n",
       "      <td>0.067455</td>\n",
       "      <td>0.069434</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>570 rows × 100 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           0         1         2         3         4         5         6   \\\n",
       "0   -0.082683 -0.151384 -0.074864  0.037651 -0.066716  0.020628 -0.011856   \n",
       "1   -0.010347  0.124159 -0.066431 -0.006549  0.084815  0.066802 -0.015166   \n",
       "2   -0.081133  0.065063  0.020470  0.089003  0.026025  0.060736 -0.083863   \n",
       "3   -0.020947 -0.121957  0.086852  0.145406 -0.006548  0.070339  0.088396   \n",
       "4   -0.007948 -0.134698 -0.004376  0.008096  0.054369 -0.095736 -0.063928   \n",
       "..        ...       ...       ...       ...       ...       ...       ...   \n",
       "565  0.092886 -0.049526  0.126560  0.222132  0.004570  0.009963  0.064991   \n",
       "566 -0.033430 -0.122807  0.129224  0.215387 -0.024991  0.105416  0.078930   \n",
       "567 -0.062368  0.007354 -0.040946  0.074390  0.186154 -0.036132  0.062150   \n",
       "568 -0.167305  0.016464 -0.089212  0.076259 -0.068883  0.175772 -0.087274   \n",
       "569  0.020393 -0.104050  0.084831 -0.064445 -0.149178  0.045278  0.170535   \n",
       "\n",
       "           7         8         9   ...        90        91        92  \\\n",
       "0   -0.013399  0.176482 -0.062232  ...  0.260981 -0.014668 -0.199746   \n",
       "1   -0.080459  0.349125  0.013353  ... -0.020182 -0.003024  0.030451   \n",
       "2   -0.036441  0.224956 -0.214574  ...  0.027686 -0.022007 -0.130763   \n",
       "3   -0.196170  0.179448 -0.215863  ...  0.096574 -0.030186 -0.172965   \n",
       "4   -0.149390  0.069077 -0.160213  ... -0.074355 -0.044783 -0.103684   \n",
       "..        ...       ...       ...  ...       ...       ...       ...   \n",
       "565 -0.123555  0.256326 -0.092037  ... -0.023978 -0.111858  0.000663   \n",
       "566 -0.059992  0.177580 -0.024562  ... -0.019837 -0.163467 -0.095012   \n",
       "567  0.035436  0.211906 -0.113501  ...  0.062830 -0.012916  0.028467   \n",
       "568 -0.135189  0.215763  0.032052  ...  0.051127 -0.065445  0.104959   \n",
       "569 -0.069695  0.126067  0.032132  ... -0.004386 -0.062179 -0.004828   \n",
       "\n",
       "           93        94        95        96        97        98        99  \n",
       "0    0.059677 -0.167955  0.148126  0.021636 -0.011042  0.135225 -0.107625  \n",
       "1    0.011098 -0.092784 -0.092140 -0.167290  0.066240  0.080714 -0.131729  \n",
       "2    0.073864 -0.040541  0.041402 -0.262502 -0.102487 -0.001430 -0.006950  \n",
       "3   -0.064311  0.070436  0.087530  0.010788 -0.023226 -0.024420 -0.023326  \n",
       "4    0.085555 -0.113270  0.023193 -0.009795 -0.042300  0.032336 -0.023733  \n",
       "..        ...       ...       ...       ...       ...       ...       ...  \n",
       "565  0.030032  0.011269  0.137639  0.026717  0.129009  0.083214  0.049911  \n",
       "566 -0.091858 -0.104963 -0.028798 -0.044007  0.096463  0.089490  0.086948  \n",
       "567 -0.000391 -0.112686  0.147133 -0.004725 -0.102680  0.070786 -0.056791  \n",
       "568  0.203136 -0.040568  0.077459  0.030946 -0.037723  0.022651 -0.017791  \n",
       "569  0.000677 -0.062164  0.117224 -0.069817  0.033966  0.067455  0.069434  \n",
       "\n",
       "[570 rows x 100 columns]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#creazione vettore per ogni recensione nel train \n",
    "\n",
    "w2v_X_train = np.zeros((len(X_train), w2v_model.vector_size))\n",
    "for i in range(len(X_train)):\n",
    "    w2v_X_train[i,:] = review_vectors(tfidf.inverse_transform(tfidf_model[i, :])[0], w2v_model.vector_size, tfidf_model[i,:].data)\n",
    "w2v_df = pd.DataFrame(w2v_X_train)\n",
    "w2v_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([-0.09799289, -0.05067216,  0.05146624,  0.11736971,  0.20772899,\n",
       "       -0.04710028,  0.09646967, -0.00158495,  0.30691639, -0.02966568,\n",
       "       -0.0238826 ,  0.11941998, -0.02537429,  0.12804612,  0.10175276,\n",
       "       -0.10420152,  0.1088097 ,  0.08596526, -0.03563311, -0.0017555 ,\n",
       "        0.02545422, -0.16633818, -0.07566929,  0.07501576,  0.11934976,\n",
       "       -0.05620475, -0.02616673,  0.05950712, -0.11073302,  0.085148  ,\n",
       "       -0.01173625,  0.01932429, -0.14097825, -0.00136308,  0.02483497,\n",
       "        0.01263816, -0.02281064,  0.0288551 , -0.05497307,  0.02782062,\n",
       "        0.11803268, -0.24067109,  0.04596454, -0.09351736, -0.11503747,\n",
       "        0.0843383 ,  0.06048436,  0.27585617,  0.04220683,  0.05171622,\n",
       "        0.01478733,  0.19195014, -0.16158716, -0.04896266,  0.10921747,\n",
       "       -0.00560302,  0.05251869,  0.12759514, -0.01391664,  0.01021231,\n",
       "       -0.00827104, -0.19851284,  0.02801869, -0.04720987,  0.19638433,\n",
       "       -0.00651141, -0.03749737,  0.00572463, -0.12428483, -0.11201406,\n",
       "        0.19590247, -0.01561656,  0.18921182, -0.02100219, -0.04468087,\n",
       "       -0.09363824, -0.03923103, -0.04283342, -0.08645181, -0.14689537,\n",
       "       -0.04782041, -0.06788214,  0.06647167, -0.05969291, -0.1078699 ,\n",
       "       -0.01331442, -0.00719705, -0.1435278 , -0.0363511 ,  0.12294968,\n",
       "        0.01363996,  0.13572659,  0.03211766,  0.07968066, -0.05498941,\n",
       "        0.0450176 , -0.0470889 ,  0.07626724,  0.02304656, -0.1575832 ])"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# creazione vettore per ogni recensione nel test\n",
    "X_test_vec = np.zeros((len(X_test), w2v_model.vector_size))\n",
    "for i in range(len(X_test)):\n",
    "    X_test_vec[i,:] = review_vectors(tfidf.inverse_transform(X_test_tf[i, :])[0], w2v_model.vector_size, X_test_tf[i,:].data)\n",
    "\n",
    "X_test_vec[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Resampled dataset shape Counter({1: 319, 0: 231})\n"
     ]
    }
   ],
   "source": [
    "from imblearn.under_sampling import TomekLinks\n",
    "\n",
    "undersampler_nn = TomekLinks(sampling_strategy=\"all\") # use tomekLinks to remove ambiguous data through nearest neighbours\n",
    "\n",
    "\n",
    "# Fit and transform the X and y data\n",
    "X_train_resampled, y_train_resampled = undersampler_nn.fit_resample(w2v_X_train, Y_train)\n",
    "\n",
    "print('Resampled dataset shape %s' % Counter(y_train_resampled))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.73      0.80      0.77        41\n",
      "           1       0.90      0.86      0.88        83\n",
      "\n",
      "    accuracy                           0.84       124\n",
      "   macro avg       0.82      0.83      0.82       124\n",
      "weighted avg       0.84      0.84      0.84       124\n",
      "\n"
     ]
    }
   ],
   "source": [
    "svm = SVC(max_iter=3000, class_weight = \"balanced\", C = 1, gamma = 1, kernel = \"poly\", probability=True)\n",
    "svm_model = svm.fit(X_train_resampled, y_train_resampled)\n",
    "predictions = svm_model.predict(X_test_vec)\n",
    "print(metrics.classification_report(Y_test, predictions))\n",
    "#recall+"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import roc_auc_score\n",
    "from scikitplot.metrics import plot_roc\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "# estrae le probabilità di ogni prediction, serve a plottare la roc-auc curve\n",
    "y_test_pred_proba = svm_model.predict_proba(X_test_vec)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.893329415221863\n"
     ]
    }
   ],
   "source": [
    "plot_roc(Y_test, svm_model.predict_proba(X_test_vec))\n",
    "plt.show()\n",
    "print(roc_auc_score(Y_test, y_test_pred_proba[:,1], average=\"macro\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "kfold = StratifiedKFold(n_splits=5, shuffle=True, random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Kernel, che determina la funzione usata per trasformare i dati in input in uno spazio ad alto numero di dimensioni;\n",
    "#C, che controlla la quantità di penalizzazione assegnata ad ev errori di classif nel processo di add del modello;\n",
    "#Gamma,  controlla l’influenza di ogni esempio del training;\n",
    "#Class weight, parametro di regolazione che viene utilizzato per gestire le classi sbilanciate.\n",
    "\n",
    "# defining parameter range\n",
    "param_grid = {'C': [0.1, 1, 10, 100],\n",
    "              'gamma': [1, 0.1, 0.01, 0.001],\n",
    "              'kernel': ['rbf','linear', 'poly'],\n",
    "              'class_weight':['balanced', None],\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 96 candidates, totalling 480 fits\n",
      "[CV 1/5] END C=0.1, class_weight=balanced, gamma=1, kernel=rbf;, score=0.864 total time=   0.0s\n",
      "[CV 2/5] END C=0.1, class_weight=balanced, gamma=1, kernel=rbf;, score=0.938 total time=   0.0s\n",
      "[CV 3/5] END C=0.1, class_weight=balanced, gamma=1, kernel=rbf;, score=0.828 total time=   0.0s\n",
      "[CV 4/5] END C=0.1, class_weight=balanced, gamma=1, kernel=rbf;, score=0.883 total time=   0.0s\n",
      "[CV 5/5] END C=0.1, class_weight=balanced, gamma=1, kernel=rbf;, score=0.889 total time=   0.0s\n",
      "[CV 1/5] END C=0.1, class_weight=balanced, gamma=1, kernel=linear;, score=0.882 total time=   0.0s\n",
      "[CV 2/5] END C=0.1, class_weight=balanced, gamma=1, kernel=linear;, score=0.937 total time=   0.0s\n",
      "[CV 3/5] END C=0.1, class_weight=balanced, gamma=1, kernel=linear;, score=0.806 total time=   0.0s\n",
      "[CV 4/5] END C=0.1, class_weight=balanced, gamma=1, kernel=linear;, score=0.902 total time=   0.0s\n",
      "[CV 5/5] END C=0.1, class_weight=balanced, gamma=1, kernel=linear;, score=0.910 total time=   0.0s\n",
      "[CV 1/5] END C=0.1, class_weight=balanced, gamma=1, kernel=poly;, score=0.860 total time=   0.0s\n",
      "[CV 2/5] END C=0.1, class_weight=balanced, gamma=1, kernel=poly;, score=0.945 total time=   0.0s\n",
      "[CV 3/5] END C=0.1, class_weight=balanced, gamma=1, kernel=poly;, score=0.823 total time=   0.0s\n",
      "[CV 4/5] END C=0.1, class_weight=balanced, gamma=1, kernel=poly;, score=0.926 total time=   0.0s\n",
      "[CV 5/5] END C=0.1, class_weight=balanced, gamma=1, kernel=poly;, score=0.905 total time=   0.0s\n",
      "[CV 1/5] END C=0.1, class_weight=balanced, gamma=0.1, kernel=rbf;, score=0.776 total time=   0.0s\n",
      "[CV 2/5] END C=0.1, class_weight=balanced, gamma=0.1, kernel=rbf;, score=0.753 total time=   0.0s\n",
      "[CV 3/5] END C=0.1, class_weight=balanced, gamma=0.1, kernel=rbf;, score=0.780 total time=   0.0s\n",
      "[CV 4/5] END C=0.1, class_weight=balanced, gamma=0.1, kernel=rbf;, score=0.762 total time=   0.0s\n",
      "[CV 5/5] END C=0.1, class_weight=balanced, gamma=0.1, kernel=rbf;, score=0.119 total time=   0.0s\n",
      "[CV 1/5] END C=0.1, class_weight=balanced, gamma=0.1, kernel=linear;, score=0.882 total time=   0.0s\n",
      "[CV 2/5] END C=0.1, class_weight=balanced, gamma=0.1, kernel=linear;, score=0.937 total time=   0.0s\n",
      "[CV 3/5] END C=0.1, class_weight=balanced, gamma=0.1, kernel=linear;, score=0.806 total time=   0.0s\n",
      "[CV 4/5] END C=0.1, class_weight=balanced, gamma=0.1, kernel=linear;, score=0.902 total time=   0.0s\n",
      "[CV 5/5] END C=0.1, class_weight=balanced, gamma=0.1, kernel=linear;, score=0.910 total time=   0.0s\n",
      "[CV 1/5] END C=0.1, class_weight=balanced, gamma=0.1, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=0.1, class_weight=balanced, gamma=0.1, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=0.1, class_weight=balanced, gamma=0.1, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=0.1, class_weight=balanced, gamma=0.1, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=0.1, class_weight=balanced, gamma=0.1, kernel=poly;, score=0.000 total time=   0.0s\n",
      "[CV 1/5] END C=0.1, class_weight=balanced, gamma=0.01, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=0.1, class_weight=balanced, gamma=0.01, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=0.1, class_weight=balanced, gamma=0.01, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=0.1, class_weight=balanced, gamma=0.01, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=0.1, class_weight=balanced, gamma=0.01, kernel=rbf;, score=0.000 total time=   0.0s\n",
      "[CV 1/5] END C=0.1, class_weight=balanced, gamma=0.01, kernel=linear;, score=0.882 total time=   0.0s\n",
      "[CV 2/5] END C=0.1, class_weight=balanced, gamma=0.01, kernel=linear;, score=0.937 total time=   0.0s\n",
      "[CV 3/5] END C=0.1, class_weight=balanced, gamma=0.01, kernel=linear;, score=0.806 total time=   0.0s\n",
      "[CV 4/5] END C=0.1, class_weight=balanced, gamma=0.01, kernel=linear;, score=0.902 total time=   0.0s\n",
      "[CV 5/5] END C=0.1, class_weight=balanced, gamma=0.01, kernel=linear;, score=0.910 total time=   0.0s\n",
      "[CV 1/5] END C=0.1, class_weight=balanced, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=0.1, class_weight=balanced, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=0.1, class_weight=balanced, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=0.1, class_weight=balanced, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=0.1, class_weight=balanced, gamma=0.01, kernel=poly;, score=0.000 total time=   0.0s\n",
      "[CV 1/5] END C=0.1, class_weight=balanced, gamma=0.001, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=0.1, class_weight=balanced, gamma=0.001, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=0.1, class_weight=balanced, gamma=0.001, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=0.1, class_weight=balanced, gamma=0.001, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=0.1, class_weight=balanced, gamma=0.001, kernel=rbf;, score=0.000 total time=   0.0s\n",
      "[CV 1/5] END C=0.1, class_weight=balanced, gamma=0.001, kernel=linear;, score=0.882 total time=   0.0s\n",
      "[CV 2/5] END C=0.1, class_weight=balanced, gamma=0.001, kernel=linear;, score=0.937 total time=   0.0s\n",
      "[CV 3/5] END C=0.1, class_weight=balanced, gamma=0.001, kernel=linear;, score=0.806 total time=   0.0s\n",
      "[CV 4/5] END C=0.1, class_weight=balanced, gamma=0.001, kernel=linear;, score=0.902 total time=   0.0s\n",
      "[CV 5/5] END C=0.1, class_weight=balanced, gamma=0.001, kernel=linear;, score=0.910 total time=   0.0s\n",
      "[CV 1/5] END C=0.1, class_weight=balanced, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=0.1, class_weight=balanced, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=0.1, class_weight=balanced, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=0.1, class_weight=balanced, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=0.1, class_weight=balanced, gamma=0.001, kernel=poly;, score=0.000 total time=   0.0s\n",
      "[CV 1/5] END C=0.1, class_weight=None, gamma=1, kernel=rbf;, score=0.857 total time=   0.0s\n",
      "[CV 2/5] END C=0.1, class_weight=None, gamma=1, kernel=rbf;, score=0.848 total time=   0.0s\n",
      "[CV 3/5] END C=0.1, class_weight=None, gamma=1, kernel=rbf;, score=0.837 total time=   0.0s\n",
      "[CV 4/5] END C=0.1, class_weight=None, gamma=1, kernel=rbf;, score=0.853 total time=   0.0s\n",
      "[CV 5/5] END C=0.1, class_weight=None, gamma=1, kernel=rbf;, score=0.818 total time=   0.0s\n",
      "[CV 1/5] END C=0.1, class_weight=None, gamma=1, kernel=linear;, score=0.874 total time=   0.0s\n",
      "[CV 2/5] END C=0.1, class_weight=None, gamma=1, kernel=linear;, score=0.899 total time=   0.0s\n",
      "[CV 3/5] END C=0.1, class_weight=None, gamma=1, kernel=linear;, score=0.874 total time=   0.0s\n",
      "[CV 4/5] END C=0.1, class_weight=None, gamma=1, kernel=linear;, score=0.912 total time=   0.0s\n",
      "[CV 5/5] END C=0.1, class_weight=None, gamma=1, kernel=linear;, score=0.881 total time=   0.0s\n",
      "[CV 1/5] END C=0.1, class_weight=None, gamma=1, kernel=poly;, score=0.749 total time=   0.0s\n",
      "[CV 2/5] END C=0.1, class_weight=None, gamma=1, kernel=poly;, score=0.753 total time=   0.0s\n",
      "[CV 3/5] END C=0.1, class_weight=None, gamma=1, kernel=poly;, score=0.753 total time=   0.0s\n",
      "[CV 4/5] END C=0.1, class_weight=None, gamma=1, kernel=poly;, score=0.757 total time=   0.0s\n",
      "[CV 5/5] END C=0.1, class_weight=None, gamma=1, kernel=poly;, score=0.746 total time=   0.0s\n",
      "[CV 1/5] END C=0.1, class_weight=None, gamma=0.1, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=0.1, class_weight=None, gamma=0.1, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=0.1, class_weight=None, gamma=0.1, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=0.1, class_weight=None, gamma=0.1, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=0.1, class_weight=None, gamma=0.1, kernel=rbf;, score=0.728 total time=   0.0s\n",
      "[CV 1/5] END C=0.1, class_weight=None, gamma=0.1, kernel=linear;, score=0.874 total time=   0.0s\n",
      "[CV 2/5] END C=0.1, class_weight=None, gamma=0.1, kernel=linear;, score=0.899 total time=   0.0s\n",
      "[CV 3/5] END C=0.1, class_weight=None, gamma=0.1, kernel=linear;, score=0.874 total time=   0.0s\n",
      "[CV 4/5] END C=0.1, class_weight=None, gamma=0.1, kernel=linear;, score=0.912 total time=   0.0s\n",
      "[CV 5/5] END C=0.1, class_weight=None, gamma=0.1, kernel=linear;, score=0.881 total time=   0.0s\n",
      "[CV 1/5] END C=0.1, class_weight=None, gamma=0.1, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=0.1, class_weight=None, gamma=0.1, kernel=poly;, score=0.736 total time=   0.0s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV 3/5] END C=0.1, class_weight=None, gamma=0.1, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=0.1, class_weight=None, gamma=0.1, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=0.1, class_weight=None, gamma=0.1, kernel=poly;, score=0.728 total time=   0.0s\n",
      "[CV 1/5] END C=0.1, class_weight=None, gamma=0.01, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=0.1, class_weight=None, gamma=0.01, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=0.1, class_weight=None, gamma=0.01, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=0.1, class_weight=None, gamma=0.01, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=0.1, class_weight=None, gamma=0.01, kernel=rbf;, score=0.728 total time=   0.0s\n",
      "[CV 1/5] END C=0.1, class_weight=None, gamma=0.01, kernel=linear;, score=0.874 total time=   0.0s\n",
      "[CV 2/5] END C=0.1, class_weight=None, gamma=0.01, kernel=linear;, score=0.899 total time=   0.0s\n",
      "[CV 3/5] END C=0.1, class_weight=None, gamma=0.01, kernel=linear;, score=0.874 total time=   0.0s\n",
      "[CV 4/5] END C=0.1, class_weight=None, gamma=0.01, kernel=linear;, score=0.912 total time=   0.0s\n",
      "[CV 5/5] END C=0.1, class_weight=None, gamma=0.01, kernel=linear;, score=0.881 total time=   0.0s\n",
      "[CV 1/5] END C=0.1, class_weight=None, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=0.1, class_weight=None, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=0.1, class_weight=None, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=0.1, class_weight=None, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=0.1, class_weight=None, gamma=0.01, kernel=poly;, score=0.728 total time=   0.0s\n",
      "[CV 1/5] END C=0.1, class_weight=None, gamma=0.001, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=0.1, class_weight=None, gamma=0.001, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=0.1, class_weight=None, gamma=0.001, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=0.1, class_weight=None, gamma=0.001, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=0.1, class_weight=None, gamma=0.001, kernel=rbf;, score=0.728 total time=   0.0s\n",
      "[CV 1/5] END C=0.1, class_weight=None, gamma=0.001, kernel=linear;, score=0.874 total time=   0.0s\n",
      "[CV 2/5] END C=0.1, class_weight=None, gamma=0.001, kernel=linear;, score=0.899 total time=   0.0s\n",
      "[CV 3/5] END C=0.1, class_weight=None, gamma=0.001, kernel=linear;, score=0.874 total time=   0.0s\n",
      "[CV 4/5] END C=0.1, class_weight=None, gamma=0.001, kernel=linear;, score=0.912 total time=   0.0s\n",
      "[CV 5/5] END C=0.1, class_weight=None, gamma=0.001, kernel=linear;, score=0.881 total time=   0.0s\n",
      "[CV 1/5] END C=0.1, class_weight=None, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=0.1, class_weight=None, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=0.1, class_weight=None, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=0.1, class_weight=None, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=0.1, class_weight=None, gamma=0.001, kernel=poly;, score=0.728 total time=   0.0s\n",
      "[CV 1/5] END C=1, class_weight=balanced, gamma=1, kernel=rbf;, score=0.882 total time=   0.0s\n",
      "[CV 2/5] END C=1, class_weight=balanced, gamma=1, kernel=rbf;, score=0.899 total time=   0.0s\n",
      "[CV 3/5] END C=1, class_weight=balanced, gamma=1, kernel=rbf;, score=0.837 total time=   0.0s\n",
      "[CV 4/5] END C=1, class_weight=balanced, gamma=1, kernel=rbf;, score=0.934 total time=   0.0s\n",
      "[CV 5/5] END C=1, class_weight=balanced, gamma=1, kernel=rbf;, score=0.896 total time=   0.0s\n",
      "[CV 1/5] END C=1, class_weight=balanced, gamma=1, kernel=linear;, score=0.866 total time=   0.0s\n",
      "[CV 2/5] END C=1, class_weight=balanced, gamma=1, kernel=linear;, score=0.912 total time=   0.0s\n",
      "[CV 3/5] END C=1, class_weight=balanced, gamma=1, kernel=linear;, score=0.812 total time=   0.0s\n",
      "[CV 4/5] END C=1, class_weight=balanced, gamma=1, kernel=linear;, score=0.918 total time=   0.0s\n",
      "[CV 5/5] END C=1, class_weight=balanced, gamma=1, kernel=linear;, score=0.896 total time=   0.0s\n",
      "[CV 1/5] END C=1, class_weight=balanced, gamma=1, kernel=poly;, score=0.905 total time=   0.0s\n",
      "[CV 2/5] END C=1, class_weight=balanced, gamma=1, kernel=poly;, score=0.906 total time=   0.0s\n",
      "[CV 3/5] END C=1, class_weight=balanced, gamma=1, kernel=poly;, score=0.868 total time=   0.0s\n",
      "[CV 4/5] END C=1, class_weight=balanced, gamma=1, kernel=poly;, score=0.944 total time=   0.0s\n",
      "[CV 5/5] END C=1, class_weight=balanced, gamma=1, kernel=poly;, score=0.902 total time=   0.0s\n",
      "[CV 1/5] END C=1, class_weight=balanced, gamma=0.1, kernel=rbf;, score=0.882 total time=   0.0s\n",
      "[CV 2/5] END C=1, class_weight=balanced, gamma=0.1, kernel=rbf;, score=0.928 total time=   0.0s\n",
      "[CV 3/5] END C=1, class_weight=balanced, gamma=0.1, kernel=rbf;, score=0.784 total time=   0.0s\n",
      "[CV 4/5] END C=1, class_weight=balanced, gamma=0.1, kernel=rbf;, score=0.900 total time=   0.0s\n",
      "[CV 5/5] END C=1, class_weight=balanced, gamma=0.1, kernel=rbf;, score=0.916 total time=   0.0s\n",
      "[CV 1/5] END C=1, class_weight=balanced, gamma=0.1, kernel=linear;, score=0.866 total time=   0.0s\n",
      "[CV 2/5] END C=1, class_weight=balanced, gamma=0.1, kernel=linear;, score=0.912 total time=   0.0s\n",
      "[CV 3/5] END C=1, class_weight=balanced, gamma=0.1, kernel=linear;, score=0.812 total time=   0.0s\n",
      "[CV 4/5] END C=1, class_weight=balanced, gamma=0.1, kernel=linear;, score=0.918 total time=   0.0s\n",
      "[CV 5/5] END C=1, class_weight=balanced, gamma=0.1, kernel=linear;, score=0.896 total time=   0.0s\n",
      "[CV 1/5] END C=1, class_weight=balanced, gamma=0.1, kernel=poly;, score=0.000 total time=   0.0s\n",
      "[CV 2/5] END C=1, class_weight=balanced, gamma=0.1, kernel=poly;, score=0.000 total time=   0.0s\n",
      "[CV 3/5] END C=1, class_weight=balanced, gamma=0.1, kernel=poly;, score=0.000 total time=   0.0s\n",
      "[CV 4/5] END C=1, class_weight=balanced, gamma=0.1, kernel=poly;, score=0.000 total time=   0.0s\n",
      "[CV 5/5] END C=1, class_weight=balanced, gamma=0.1, kernel=poly;, score=0.728 total time=   0.0s\n",
      "[CV 1/5] END C=1, class_weight=balanced, gamma=0.01, kernel=rbf;, score=0.222 total time=   0.0s\n",
      "[CV 2/5] END C=1, class_weight=balanced, gamma=0.01, kernel=rbf;, score=0.247 total time=   0.0s\n",
      "[CV 3/5] END C=1, class_weight=balanced, gamma=0.01, kernel=rbf;, score=0.338 total time=   0.0s\n",
      "[CV 4/5] END C=1, class_weight=balanced, gamma=0.01, kernel=rbf;, score=0.145 total time=   0.0s\n",
      "[CV 5/5] END C=1, class_weight=balanced, gamma=0.01, kernel=rbf;, score=0.803 total time=   0.0s\n",
      "[CV 1/5] END C=1, class_weight=balanced, gamma=0.01, kernel=linear;, score=0.866 total time=   0.0s\n",
      "[CV 2/5] END C=1, class_weight=balanced, gamma=0.01, kernel=linear;, score=0.912 total time=   0.0s\n",
      "[CV 3/5] END C=1, class_weight=balanced, gamma=0.01, kernel=linear;, score=0.812 total time=   0.0s\n",
      "[CV 4/5] END C=1, class_weight=balanced, gamma=0.01, kernel=linear;, score=0.918 total time=   0.0s\n",
      "[CV 5/5] END C=1, class_weight=balanced, gamma=0.01, kernel=linear;, score=0.896 total time=   0.0s\n",
      "[CV 1/5] END C=1, class_weight=balanced, gamma=0.01, kernel=poly;, score=0.000 total time=   0.0s\n",
      "[CV 2/5] END C=1, class_weight=balanced, gamma=0.01, kernel=poly;, score=0.000 total time=   0.0s\n",
      "[CV 3/5] END C=1, class_weight=balanced, gamma=0.01, kernel=poly;, score=0.000 total time=   0.0s\n",
      "[CV 4/5] END C=1, class_weight=balanced, gamma=0.01, kernel=poly;, score=0.000 total time=   0.0s\n",
      "[CV 5/5] END C=1, class_weight=balanced, gamma=0.01, kernel=poly;, score=0.728 total time=   0.0s\n",
      "[CV 1/5] END C=1, class_weight=balanced, gamma=0.001, kernel=rbf;, score=0.000 total time=   0.0s\n",
      "[CV 2/5] END C=1, class_weight=balanced, gamma=0.001, kernel=rbf;, score=0.000 total time=   0.0s\n",
      "[CV 3/5] END C=1, class_weight=balanced, gamma=0.001, kernel=rbf;, score=0.000 total time=   0.0s\n",
      "[CV 4/5] END C=1, class_weight=balanced, gamma=0.001, kernel=rbf;, score=0.000 total time=   0.0s\n",
      "[CV 5/5] END C=1, class_weight=balanced, gamma=0.001, kernel=rbf;, score=0.728 total time=   0.0s\n",
      "[CV 1/5] END C=1, class_weight=balanced, gamma=0.001, kernel=linear;, score=0.866 total time=   0.0s\n",
      "[CV 2/5] END C=1, class_weight=balanced, gamma=0.001, kernel=linear;, score=0.912 total time=   0.0s\n",
      "[CV 3/5] END C=1, class_weight=balanced, gamma=0.001, kernel=linear;, score=0.812 total time=   0.0s\n",
      "[CV 4/5] END C=1, class_weight=balanced, gamma=0.001, kernel=linear;, score=0.918 total time=   0.0s\n",
      "[CV 5/5] END C=1, class_weight=balanced, gamma=0.001, kernel=linear;, score=0.896 total time=   0.0s\n",
      "[CV 1/5] END C=1, class_weight=balanced, gamma=0.001, kernel=poly;, score=0.000 total time=   0.0s\n",
      "[CV 2/5] END C=1, class_weight=balanced, gamma=0.001, kernel=poly;, score=0.000 total time=   0.0s\n",
      "[CV 3/5] END C=1, class_weight=balanced, gamma=0.001, kernel=poly;, score=0.000 total time=   0.0s\n",
      "[CV 4/5] END C=1, class_weight=balanced, gamma=0.001, kernel=poly;, score=0.000 total time=   0.0s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV 5/5] END C=1, class_weight=balanced, gamma=0.001, kernel=poly;, score=0.728 total time=   0.0s\n",
      "[CV 1/5] END C=1, class_weight=None, gamma=1, kernel=rbf;, score=0.875 total time=   0.0s\n",
      "[CV 2/5] END C=1, class_weight=None, gamma=1, kernel=rbf;, score=0.908 total time=   0.0s\n",
      "[CV 3/5] END C=1, class_weight=None, gamma=1, kernel=rbf;, score=0.848 total time=   0.0s\n",
      "[CV 4/5] END C=1, class_weight=None, gamma=1, kernel=rbf;, score=0.922 total time=   0.0s\n",
      "[CV 5/5] END C=1, class_weight=None, gamma=1, kernel=rbf;, score=0.896 total time=   0.0s\n",
      "[CV 1/5] END C=1, class_weight=None, gamma=1, kernel=linear;, score=0.877 total time=   0.0s\n",
      "[CV 2/5] END C=1, class_weight=None, gamma=1, kernel=linear;, score=0.885 total time=   0.0s\n",
      "[CV 3/5] END C=1, class_weight=None, gamma=1, kernel=linear;, score=0.831 total time=   0.0s\n",
      "[CV 4/5] END C=1, class_weight=None, gamma=1, kernel=linear;, score=0.929 total time=   0.0s\n",
      "[CV 5/5] END C=1, class_weight=None, gamma=1, kernel=linear;, score=0.891 total time=   0.0s\n",
      "[CV 1/5] END C=1, class_weight=None, gamma=1, kernel=poly;, score=0.901 total time=   0.0s\n",
      "[CV 2/5] END C=1, class_weight=None, gamma=1, kernel=poly;, score=0.916 total time=   0.0s\n",
      "[CV 3/5] END C=1, class_weight=None, gamma=1, kernel=poly;, score=0.870 total time=   0.0s\n",
      "[CV 4/5] END C=1, class_weight=None, gamma=1, kernel=poly;, score=0.946 total time=   0.0s\n",
      "[CV 5/5] END C=1, class_weight=None, gamma=1, kernel=poly;, score=0.896 total time=   0.0s\n",
      "[CV 1/5] END C=1, class_weight=None, gamma=0.1, kernel=rbf;, score=0.868 total time=   0.0s\n",
      "[CV 2/5] END C=1, class_weight=None, gamma=0.1, kernel=rbf;, score=0.905 total time=   0.0s\n",
      "[CV 3/5] END C=1, class_weight=None, gamma=0.1, kernel=rbf;, score=0.857 total time=   0.0s\n",
      "[CV 4/5] END C=1, class_weight=None, gamma=0.1, kernel=rbf;, score=0.919 total time=   0.0s\n",
      "[CV 5/5] END C=1, class_weight=None, gamma=0.1, kernel=rbf;, score=0.892 total time=   0.0s\n",
      "[CV 1/5] END C=1, class_weight=None, gamma=0.1, kernel=linear;, score=0.877 total time=   0.0s\n",
      "[CV 2/5] END C=1, class_weight=None, gamma=0.1, kernel=linear;, score=0.885 total time=   0.0s\n",
      "[CV 3/5] END C=1, class_weight=None, gamma=0.1, kernel=linear;, score=0.831 total time=   0.0s\n",
      "[CV 4/5] END C=1, class_weight=None, gamma=0.1, kernel=linear;, score=0.929 total time=   0.0s\n",
      "[CV 5/5] END C=1, class_weight=None, gamma=0.1, kernel=linear;, score=0.891 total time=   0.0s\n",
      "[CV 1/5] END C=1, class_weight=None, gamma=0.1, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=1, class_weight=None, gamma=0.1, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=1, class_weight=None, gamma=0.1, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=1, class_weight=None, gamma=0.1, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=1, class_weight=None, gamma=0.1, kernel=poly;, score=0.728 total time=   0.0s\n",
      "[CV 1/5] END C=1, class_weight=None, gamma=0.01, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=1, class_weight=None, gamma=0.01, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=1, class_weight=None, gamma=0.01, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=1, class_weight=None, gamma=0.01, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=1, class_weight=None, gamma=0.01, kernel=rbf;, score=0.728 total time=   0.0s\n",
      "[CV 1/5] END C=1, class_weight=None, gamma=0.01, kernel=linear;, score=0.877 total time=   0.0s\n",
      "[CV 2/5] END C=1, class_weight=None, gamma=0.01, kernel=linear;, score=0.885 total time=   0.0s\n",
      "[CV 3/5] END C=1, class_weight=None, gamma=0.01, kernel=linear;, score=0.831 total time=   0.0s\n",
      "[CV 4/5] END C=1, class_weight=None, gamma=0.01, kernel=linear;, score=0.929 total time=   0.0s\n",
      "[CV 5/5] END C=1, class_weight=None, gamma=0.01, kernel=linear;, score=0.891 total time=   0.0s\n",
      "[CV 1/5] END C=1, class_weight=None, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=1, class_weight=None, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=1, class_weight=None, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=1, class_weight=None, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=1, class_weight=None, gamma=0.01, kernel=poly;, score=0.728 total time=   0.0s\n",
      "[CV 1/5] END C=1, class_weight=None, gamma=0.001, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=1, class_weight=None, gamma=0.001, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=1, class_weight=None, gamma=0.001, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=1, class_weight=None, gamma=0.001, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=1, class_weight=None, gamma=0.001, kernel=rbf;, score=0.728 total time=   0.0s\n",
      "[CV 1/5] END C=1, class_weight=None, gamma=0.001, kernel=linear;, score=0.877 total time=   0.0s\n",
      "[CV 2/5] END C=1, class_weight=None, gamma=0.001, kernel=linear;, score=0.885 total time=   0.0s\n",
      "[CV 3/5] END C=1, class_weight=None, gamma=0.001, kernel=linear;, score=0.831 total time=   0.0s\n",
      "[CV 4/5] END C=1, class_weight=None, gamma=0.001, kernel=linear;, score=0.929 total time=   0.0s\n",
      "[CV 5/5] END C=1, class_weight=None, gamma=0.001, kernel=linear;, score=0.891 total time=   0.0s\n",
      "[CV 1/5] END C=1, class_weight=None, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=1, class_weight=None, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=1, class_weight=None, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=1, class_weight=None, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=1, class_weight=None, gamma=0.001, kernel=poly;, score=0.728 total time=   0.0s\n",
      "[CV 1/5] END C=10, class_weight=balanced, gamma=1, kernel=rbf;, score=0.882 total time=   0.0s\n",
      "[CV 2/5] END C=10, class_weight=balanced, gamma=1, kernel=rbf;, score=0.889 total time=   0.0s\n",
      "[CV 3/5] END C=10, class_weight=balanced, gamma=1, kernel=rbf;, score=0.873 total time=   0.0s\n",
      "[CV 4/5] END C=10, class_weight=balanced, gamma=1, kernel=rbf;, score=0.935 total time=   0.0s\n",
      "[CV 5/5] END C=10, class_weight=balanced, gamma=1, kernel=rbf;, score=0.887 total time=   0.0s\n",
      "[CV 1/5] END C=10, class_weight=balanced, gamma=1, kernel=linear;, score=0.836 total time=   0.0s\n",
      "[CV 2/5] END C=10, class_weight=balanced, gamma=1, kernel=linear;, score=0.866 total time=   0.0s\n",
      "[CV 3/5] END C=10, class_weight=balanced, gamma=1, kernel=linear;, score=0.846 total time=   0.0s\n",
      "[CV 4/5] END C=10, class_weight=balanced, gamma=1, kernel=linear;, score=0.894 total time=   0.0s\n",
      "[CV 5/5] END C=10, class_weight=balanced, gamma=1, kernel=linear;, score=0.824 total time=   0.0s\n",
      "[CV 1/5] END C=10, class_weight=balanced, gamma=1, kernel=poly;, score=0.891 total time=   0.0s\n",
      "[CV 2/5] END C=10, class_weight=balanced, gamma=1, kernel=poly;, score=0.889 total time=   0.0s\n",
      "[CV 3/5] END C=10, class_weight=balanced, gamma=1, kernel=poly;, score=0.882 total time=   0.0s\n",
      "[CV 4/5] END C=10, class_weight=balanced, gamma=1, kernel=poly;, score=0.945 total time=   0.0s\n",
      "[CV 5/5] END C=10, class_weight=balanced, gamma=1, kernel=poly;, score=0.881 total time=   0.0s\n",
      "[CV 1/5] END C=10, class_weight=balanced, gamma=0.1, kernel=rbf;, score=0.857 total time=   0.0s\n",
      "[CV 2/5] END C=10, class_weight=balanced, gamma=0.1, kernel=rbf;, score=0.905 total time=   0.0s\n",
      "[CV 3/5] END C=10, class_weight=balanced, gamma=0.1, kernel=rbf;, score=0.812 total time=   0.0s\n",
      "[CV 4/5] END C=10, class_weight=balanced, gamma=0.1, kernel=rbf;, score=0.917 total time=   0.0s\n",
      "[CV 5/5] END C=10, class_weight=balanced, gamma=0.1, kernel=rbf;, score=0.889 total time=   0.0s\n",
      "[CV 1/5] END C=10, class_weight=balanced, gamma=0.1, kernel=linear;, score=0.836 total time=   0.0s\n",
      "[CV 2/5] END C=10, class_weight=balanced, gamma=0.1, kernel=linear;, score=0.866 total time=   0.0s\n",
      "[CV 3/5] END C=10, class_weight=balanced, gamma=0.1, kernel=linear;, score=0.846 total time=   0.0s\n",
      "[CV 4/5] END C=10, class_weight=balanced, gamma=0.1, kernel=linear;, score=0.894 total time=   0.0s\n",
      "[CV 5/5] END C=10, class_weight=balanced, gamma=0.1, kernel=linear;, score=0.824 total time=   0.0s\n",
      "[CV 1/5] END C=10, class_weight=balanced, gamma=0.1, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=10, class_weight=balanced, gamma=0.1, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=10, class_weight=balanced, gamma=0.1, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=10, class_weight=balanced, gamma=0.1, kernel=poly;, score=0.736 total time=   0.0s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV 5/5] END C=10, class_weight=balanced, gamma=0.1, kernel=poly;, score=0.728 total time=   0.0s\n",
      "[CV 1/5] END C=10, class_weight=balanced, gamma=0.01, kernel=rbf;, score=0.889 total time=   0.0s\n",
      "[CV 2/5] END C=10, class_weight=balanced, gamma=0.01, kernel=rbf;, score=0.929 total time=   0.0s\n",
      "[CV 3/5] END C=10, class_weight=balanced, gamma=0.01, kernel=rbf;, score=0.784 total time=   0.0s\n",
      "[CV 4/5] END C=10, class_weight=balanced, gamma=0.01, kernel=rbf;, score=0.900 total time=   0.0s\n",
      "[CV 5/5] END C=10, class_weight=balanced, gamma=0.01, kernel=rbf;, score=0.909 total time=   0.0s\n",
      "[CV 1/5] END C=10, class_weight=balanced, gamma=0.01, kernel=linear;, score=0.836 total time=   0.0s\n",
      "[CV 2/5] END C=10, class_weight=balanced, gamma=0.01, kernel=linear;, score=0.866 total time=   0.0s\n",
      "[CV 3/5] END C=10, class_weight=balanced, gamma=0.01, kernel=linear;, score=0.846 total time=   0.0s\n",
      "[CV 4/5] END C=10, class_weight=balanced, gamma=0.01, kernel=linear;, score=0.894 total time=   0.0s\n",
      "[CV 5/5] END C=10, class_weight=balanced, gamma=0.01, kernel=linear;, score=0.824 total time=   0.0s\n",
      "[CV 1/5] END C=10, class_weight=balanced, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=10, class_weight=balanced, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=10, class_weight=balanced, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=10, class_weight=balanced, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=10, class_weight=balanced, gamma=0.01, kernel=poly;, score=0.728 total time=   0.0s\n",
      "[CV 1/5] END C=10, class_weight=balanced, gamma=0.001, kernel=rbf;, score=0.821 total time=   0.0s\n",
      "[CV 2/5] END C=10, class_weight=balanced, gamma=0.001, kernel=rbf;, score=0.800 total time=   0.0s\n",
      "[CV 3/5] END C=10, class_weight=balanced, gamma=0.001, kernel=rbf;, score=0.831 total time=   0.0s\n",
      "[CV 4/5] END C=10, class_weight=balanced, gamma=0.001, kernel=rbf;, score=0.815 total time=   0.0s\n",
      "[CV 5/5] END C=10, class_weight=balanced, gamma=0.001, kernel=rbf;, score=0.803 total time=   0.0s\n",
      "[CV 1/5] END C=10, class_weight=balanced, gamma=0.001, kernel=linear;, score=0.836 total time=   0.0s\n",
      "[CV 2/5] END C=10, class_weight=balanced, gamma=0.001, kernel=linear;, score=0.866 total time=   0.0s\n",
      "[CV 3/5] END C=10, class_weight=balanced, gamma=0.001, kernel=linear;, score=0.846 total time=   0.0s\n",
      "[CV 4/5] END C=10, class_weight=balanced, gamma=0.001, kernel=linear;, score=0.894 total time=   0.0s\n",
      "[CV 5/5] END C=10, class_weight=balanced, gamma=0.001, kernel=linear;, score=0.824 total time=   0.0s\n",
      "[CV 1/5] END C=10, class_weight=balanced, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=10, class_weight=balanced, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=10, class_weight=balanced, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=10, class_weight=balanced, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=10, class_weight=balanced, gamma=0.001, kernel=poly;, score=0.728 total time=   0.0s\n",
      "[CV 1/5] END C=10, class_weight=None, gamma=1, kernel=rbf;, score=0.882 total time=   0.0s\n",
      "[CV 2/5] END C=10, class_weight=None, gamma=1, kernel=rbf;, score=0.882 total time=   0.0s\n",
      "[CV 3/5] END C=10, class_weight=None, gamma=1, kernel=rbf;, score=0.873 total time=   0.0s\n",
      "[CV 4/5] END C=10, class_weight=None, gamma=1, kernel=rbf;, score=0.921 total time=   0.0s\n",
      "[CV 5/5] END C=10, class_weight=None, gamma=1, kernel=rbf;, score=0.887 total time=   0.0s\n",
      "[CV 1/5] END C=10, class_weight=None, gamma=1, kernel=linear;, score=0.855 total time=   0.0s\n",
      "[CV 2/5] END C=10, class_weight=None, gamma=1, kernel=linear;, score=0.868 total time=   0.0s\n",
      "[CV 3/5] END C=10, class_weight=None, gamma=1, kernel=linear;, score=0.822 total time=   0.0s\n",
      "[CV 4/5] END C=10, class_weight=None, gamma=1, kernel=linear;, score=0.889 total time=   0.0s\n",
      "[CV 5/5] END C=10, class_weight=None, gamma=1, kernel=linear;, score=0.830 total time=   0.0s\n",
      "[CV 1/5] END C=10, class_weight=None, gamma=1, kernel=poly;, score=0.901 total time=   0.0s\n",
      "[CV 2/5] END C=10, class_weight=None, gamma=1, kernel=poly;, score=0.906 total time=   0.0s\n",
      "[CV 3/5] END C=10, class_weight=None, gamma=1, kernel=poly;, score=0.891 total time=   0.0s\n",
      "[CV 4/5] END C=10, class_weight=None, gamma=1, kernel=poly;, score=0.946 total time=   0.0s\n",
      "[CV 5/5] END C=10, class_weight=None, gamma=1, kernel=poly;, score=0.881 total time=   0.0s\n",
      "[CV 1/5] END C=10, class_weight=None, gamma=0.1, kernel=rbf;, score=0.875 total time=   0.0s\n",
      "[CV 2/5] END C=10, class_weight=None, gamma=0.1, kernel=rbf;, score=0.908 total time=   0.0s\n",
      "[CV 3/5] END C=10, class_weight=None, gamma=0.1, kernel=rbf;, score=0.857 total time=   0.0s\n",
      "[CV 4/5] END C=10, class_weight=None, gamma=0.1, kernel=rbf;, score=0.929 total time=   0.0s\n",
      "[CV 5/5] END C=10, class_weight=None, gamma=0.1, kernel=rbf;, score=0.891 total time=   0.0s\n",
      "[CV 1/5] END C=10, class_weight=None, gamma=0.1, kernel=linear;, score=0.855 total time=   0.0s\n",
      "[CV 2/5] END C=10, class_weight=None, gamma=0.1, kernel=linear;, score=0.868 total time=   0.0s\n",
      "[CV 3/5] END C=10, class_weight=None, gamma=0.1, kernel=linear;, score=0.822 total time=   0.0s\n",
      "[CV 4/5] END C=10, class_weight=None, gamma=0.1, kernel=linear;, score=0.889 total time=   0.0s\n",
      "[CV 5/5] END C=10, class_weight=None, gamma=0.1, kernel=linear;, score=0.830 total time=   0.0s\n",
      "[CV 1/5] END C=10, class_weight=None, gamma=0.1, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=10, class_weight=None, gamma=0.1, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=10, class_weight=None, gamma=0.1, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=10, class_weight=None, gamma=0.1, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=10, class_weight=None, gamma=0.1, kernel=poly;, score=0.728 total time=   0.0s\n",
      "[CV 1/5] END C=10, class_weight=None, gamma=0.01, kernel=rbf;, score=0.875 total time=   0.0s\n",
      "[CV 2/5] END C=10, class_weight=None, gamma=0.01, kernel=rbf;, score=0.905 total time=   0.0s\n",
      "[CV 3/5] END C=10, class_weight=None, gamma=0.01, kernel=rbf;, score=0.857 total time=   0.0s\n",
      "[CV 4/5] END C=10, class_weight=None, gamma=0.01, kernel=rbf;, score=0.912 total time=   0.0s\n",
      "[CV 5/5] END C=10, class_weight=None, gamma=0.01, kernel=rbf;, score=0.886 total time=   0.0s\n",
      "[CV 1/5] END C=10, class_weight=None, gamma=0.01, kernel=linear;, score=0.855 total time=   0.0s\n",
      "[CV 2/5] END C=10, class_weight=None, gamma=0.01, kernel=linear;, score=0.868 total time=   0.0s\n",
      "[CV 3/5] END C=10, class_weight=None, gamma=0.01, kernel=linear;, score=0.822 total time=   0.0s\n",
      "[CV 4/5] END C=10, class_weight=None, gamma=0.01, kernel=linear;, score=0.889 total time=   0.0s\n",
      "[CV 5/5] END C=10, class_weight=None, gamma=0.01, kernel=linear;, score=0.830 total time=   0.0s\n",
      "[CV 1/5] END C=10, class_weight=None, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=10, class_weight=None, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=10, class_weight=None, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=10, class_weight=None, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=10, class_weight=None, gamma=0.01, kernel=poly;, score=0.728 total time=   0.0s\n",
      "[CV 1/5] END C=10, class_weight=None, gamma=0.001, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=10, class_weight=None, gamma=0.001, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=10, class_weight=None, gamma=0.001, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=10, class_weight=None, gamma=0.001, kernel=rbf;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=10, class_weight=None, gamma=0.001, kernel=rbf;, score=0.728 total time=   0.0s\n",
      "[CV 1/5] END C=10, class_weight=None, gamma=0.001, kernel=linear;, score=0.855 total time=   0.0s\n",
      "[CV 2/5] END C=10, class_weight=None, gamma=0.001, kernel=linear;, score=0.868 total time=   0.0s\n",
      "[CV 3/5] END C=10, class_weight=None, gamma=0.001, kernel=linear;, score=0.822 total time=   0.0s\n",
      "[CV 4/5] END C=10, class_weight=None, gamma=0.001, kernel=linear;, score=0.889 total time=   0.0s\n",
      "[CV 5/5] END C=10, class_weight=None, gamma=0.001, kernel=linear;, score=0.830 total time=   0.0s\n",
      "[CV 1/5] END C=10, class_weight=None, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=10, class_weight=None, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=10, class_weight=None, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=10, class_weight=None, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=10, class_weight=None, gamma=0.001, kernel=poly;, score=0.728 total time=   0.0s\n",
      "[CV 1/5] END C=100, class_weight=balanced, gamma=1, kernel=rbf;, score=0.882 total time=   0.0s\n",
      "[CV 2/5] END C=100, class_weight=balanced, gamma=1, kernel=rbf;, score=0.889 total time=   0.0s\n",
      "[CV 3/5] END C=100, class_weight=balanced, gamma=1, kernel=rbf;, score=0.873 total time=   0.0s\n",
      "[CV 4/5] END C=100, class_weight=balanced, gamma=1, kernel=rbf;, score=0.935 total time=   0.0s\n",
      "[CV 5/5] END C=100, class_weight=balanced, gamma=1, kernel=rbf;, score=0.887 total time=   0.0s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV 1/5] END C=100, class_weight=balanced, gamma=1, kernel=linear;, score=0.850 total time=   0.0s\n",
      "[CV 2/5] END C=100, class_weight=balanced, gamma=1, kernel=linear;, score=0.843 total time=   0.0s\n",
      "[CV 3/5] END C=100, class_weight=balanced, gamma=1, kernel=linear;, score=0.835 total time=   0.0s\n",
      "[CV 4/5] END C=100, class_weight=balanced, gamma=1, kernel=linear;, score=0.889 total time=   0.0s\n",
      "[CV 5/5] END C=100, class_weight=balanced, gamma=1, kernel=linear;, score=0.773 total time=   0.0s\n",
      "[CV 1/5] END C=100, class_weight=balanced, gamma=1, kernel=poly;, score=0.894 total time=   0.0s\n",
      "[CV 2/5] END C=100, class_weight=balanced, gamma=1, kernel=poly;, score=0.862 total time=   0.0s\n",
      "[CV 3/5] END C=100, class_weight=balanced, gamma=1, kernel=poly;, score=0.878 total time=   0.0s\n",
      "[CV 4/5] END C=100, class_weight=balanced, gamma=1, kernel=poly;, score=0.937 total time=   0.0s\n",
      "[CV 5/5] END C=100, class_weight=balanced, gamma=1, kernel=poly;, score=0.861 total time=   0.0s\n",
      "[CV 1/5] END C=100, class_weight=balanced, gamma=0.1, kernel=rbf;, score=0.852 total time=   0.0s\n",
      "[CV 2/5] END C=100, class_weight=balanced, gamma=0.1, kernel=rbf;, score=0.880 total time=   0.0s\n",
      "[CV 3/5] END C=100, class_weight=balanced, gamma=0.1, kernel=rbf;, score=0.868 total time=   0.0s\n",
      "[CV 4/5] END C=100, class_weight=balanced, gamma=0.1, kernel=rbf;, score=0.906 total time=   0.0s\n",
      "[CV 5/5] END C=100, class_weight=balanced, gamma=0.1, kernel=rbf;, score=0.830 total time=   0.0s\n",
      "[CV 1/5] END C=100, class_weight=balanced, gamma=0.1, kernel=linear;, score=0.850 total time=   0.0s\n",
      "[CV 2/5] END C=100, class_weight=balanced, gamma=0.1, kernel=linear;, score=0.843 total time=   0.0s\n",
      "[CV 3/5] END C=100, class_weight=balanced, gamma=0.1, kernel=linear;, score=0.835 total time=   0.0s\n",
      "[CV 4/5] END C=100, class_weight=balanced, gamma=0.1, kernel=linear;, score=0.889 total time=   0.0s\n",
      "[CV 5/5] END C=100, class_weight=balanced, gamma=0.1, kernel=linear;, score=0.773 total time=   0.0s\n",
      "[CV 1/5] END C=100, class_weight=balanced, gamma=0.1, kernel=poly;, score=0.860 total time=   0.0s\n",
      "[CV 2/5] END C=100, class_weight=balanced, gamma=0.1, kernel=poly;, score=0.945 total time=   0.0s\n",
      "[CV 3/5] END C=100, class_weight=balanced, gamma=0.1, kernel=poly;, score=0.823 total time=   0.0s\n",
      "[CV 4/5] END C=100, class_weight=balanced, gamma=0.1, kernel=poly;, score=0.926 total time=   0.0s\n",
      "[CV 5/5] END C=100, class_weight=balanced, gamma=0.1, kernel=poly;, score=0.905 total time=   0.0s\n",
      "[CV 1/5] END C=100, class_weight=balanced, gamma=0.01, kernel=rbf;, score=0.864 total time=   0.0s\n",
      "[CV 2/5] END C=100, class_weight=balanced, gamma=0.01, kernel=rbf;, score=0.894 total time=   0.0s\n",
      "[CV 3/5] END C=100, class_weight=balanced, gamma=0.01, kernel=rbf;, score=0.812 total time=   0.0s\n",
      "[CV 4/5] END C=100, class_weight=balanced, gamma=0.01, kernel=rbf;, score=0.917 total time=   0.0s\n",
      "[CV 5/5] END C=100, class_weight=balanced, gamma=0.01, kernel=rbf;, score=0.874 total time=   0.0s\n",
      "[CV 1/5] END C=100, class_weight=balanced, gamma=0.01, kernel=linear;, score=0.850 total time=   0.0s\n",
      "[CV 2/5] END C=100, class_weight=balanced, gamma=0.01, kernel=linear;, score=0.843 total time=   0.0s\n",
      "[CV 3/5] END C=100, class_weight=balanced, gamma=0.01, kernel=linear;, score=0.835 total time=   0.0s\n",
      "[CV 4/5] END C=100, class_weight=balanced, gamma=0.01, kernel=linear;, score=0.889 total time=   0.0s\n",
      "[CV 5/5] END C=100, class_weight=balanced, gamma=0.01, kernel=linear;, score=0.773 total time=   0.0s\n",
      "[CV 1/5] END C=100, class_weight=balanced, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=100, class_weight=balanced, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=100, class_weight=balanced, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=100, class_weight=balanced, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=100, class_weight=balanced, gamma=0.01, kernel=poly;, score=0.728 total time=   0.0s\n",
      "[CV 1/5] END C=100, class_weight=balanced, gamma=0.001, kernel=rbf;, score=0.889 total time=   0.0s\n",
      "[CV 2/5] END C=100, class_weight=balanced, gamma=0.001, kernel=rbf;, score=0.929 total time=   0.0s\n",
      "[CV 3/5] END C=100, class_weight=balanced, gamma=0.001, kernel=rbf;, score=0.784 total time=   0.0s\n",
      "[CV 4/5] END C=100, class_weight=balanced, gamma=0.001, kernel=rbf;, score=0.900 total time=   0.0s\n",
      "[CV 5/5] END C=100, class_weight=balanced, gamma=0.001, kernel=rbf;, score=0.909 total time=   0.0s\n",
      "[CV 1/5] END C=100, class_weight=balanced, gamma=0.001, kernel=linear;, score=0.850 total time=   0.0s\n",
      "[CV 2/5] END C=100, class_weight=balanced, gamma=0.001, kernel=linear;, score=0.843 total time=   0.0s\n",
      "[CV 3/5] END C=100, class_weight=balanced, gamma=0.001, kernel=linear;, score=0.835 total time=   0.0s\n",
      "[CV 4/5] END C=100, class_weight=balanced, gamma=0.001, kernel=linear;, score=0.889 total time=   0.0s\n",
      "[CV 5/5] END C=100, class_weight=balanced, gamma=0.001, kernel=linear;, score=0.773 total time=   0.0s\n",
      "[CV 1/5] END C=100, class_weight=balanced, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=100, class_weight=balanced, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=100, class_weight=balanced, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=100, class_weight=balanced, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=100, class_weight=balanced, gamma=0.001, kernel=poly;, score=0.728 total time=   0.0s\n",
      "[CV 1/5] END C=100, class_weight=None, gamma=1, kernel=rbf;, score=0.882 total time=   0.0s\n",
      "[CV 2/5] END C=100, class_weight=None, gamma=1, kernel=rbf;, score=0.882 total time=   0.0s\n",
      "[CV 3/5] END C=100, class_weight=None, gamma=1, kernel=rbf;, score=0.873 total time=   0.0s\n",
      "[CV 4/5] END C=100, class_weight=None, gamma=1, kernel=rbf;, score=0.921 total time=   0.0s\n",
      "[CV 5/5] END C=100, class_weight=None, gamma=1, kernel=rbf;, score=0.887 total time=   0.0s\n",
      "[CV 1/5] END C=100, class_weight=None, gamma=1, kernel=linear;, score=0.866 total time=   0.0s\n",
      "[CV 2/5] END C=100, class_weight=None, gamma=1, kernel=linear;, score=0.864 total time=   0.0s\n",
      "[CV 3/5] END C=100, class_weight=None, gamma=1, kernel=linear;, score=0.806 total time=   0.0s\n",
      "[CV 4/5] END C=100, class_weight=None, gamma=1, kernel=linear;, score=0.871 total time=   0.0s\n",
      "[CV 5/5] END C=100, class_weight=None, gamma=1, kernel=linear;, score=0.785 total time=   0.0s\n",
      "[CV 1/5] END C=100, class_weight=None, gamma=1, kernel=poly;, score=0.901 total time=   0.0s\n",
      "[CV 2/5] END C=100, class_weight=None, gamma=1, kernel=poly;, score=0.906 total time=   0.0s\n",
      "[CV 3/5] END C=100, class_weight=None, gamma=1, kernel=poly;, score=0.891 total time=   0.0s\n",
      "[CV 4/5] END C=100, class_weight=None, gamma=1, kernel=poly;, score=0.909 total time=   0.0s\n",
      "[CV 5/5] END C=100, class_weight=None, gamma=1, kernel=poly;, score=0.861 total time=   0.0s\n",
      "[CV 1/5] END C=100, class_weight=None, gamma=0.1, kernel=rbf;, score=0.862 total time=   0.0s\n",
      "[CV 2/5] END C=100, class_weight=None, gamma=0.1, kernel=rbf;, score=0.866 total time=   0.0s\n",
      "[CV 3/5] END C=100, class_weight=None, gamma=0.1, kernel=rbf;, score=0.862 total time=   0.0s\n",
      "[CV 4/5] END C=100, class_weight=None, gamma=0.1, kernel=rbf;, score=0.906 total time=   0.0s\n",
      "[CV 5/5] END C=100, class_weight=None, gamma=0.1, kernel=rbf;, score=0.838 total time=   0.0s\n",
      "[CV 1/5] END C=100, class_weight=None, gamma=0.1, kernel=linear;, score=0.866 total time=   0.0s\n",
      "[CV 2/5] END C=100, class_weight=None, gamma=0.1, kernel=linear;, score=0.864 total time=   0.0s\n",
      "[CV 3/5] END C=100, class_weight=None, gamma=0.1, kernel=linear;, score=0.806 total time=   0.0s\n",
      "[CV 4/5] END C=100, class_weight=None, gamma=0.1, kernel=linear;, score=0.871 total time=   0.0s\n",
      "[CV 5/5] END C=100, class_weight=None, gamma=0.1, kernel=linear;, score=0.785 total time=   0.0s\n",
      "[CV 1/5] END C=100, class_weight=None, gamma=0.1, kernel=poly;, score=0.749 total time=   0.0s\n",
      "[CV 2/5] END C=100, class_weight=None, gamma=0.1, kernel=poly;, score=0.753 total time=   0.0s\n",
      "[CV 3/5] END C=100, class_weight=None, gamma=0.1, kernel=poly;, score=0.753 total time=   0.0s\n",
      "[CV 4/5] END C=100, class_weight=None, gamma=0.1, kernel=poly;, score=0.757 total time=   0.0s\n",
      "[CV 5/5] END C=100, class_weight=None, gamma=0.1, kernel=poly;, score=0.746 total time=   0.0s\n",
      "[CV 1/5] END C=100, class_weight=None, gamma=0.01, kernel=rbf;, score=0.857 total time=   0.0s\n",
      "[CV 2/5] END C=100, class_weight=None, gamma=0.01, kernel=rbf;, score=0.908 total time=   0.0s\n",
      "[CV 3/5] END C=100, class_weight=None, gamma=0.01, kernel=rbf;, score=0.848 total time=   0.0s\n",
      "[CV 4/5] END C=100, class_weight=None, gamma=0.01, kernel=rbf;, score=0.921 total time=   0.0s\n",
      "[CV 5/5] END C=100, class_weight=None, gamma=0.01, kernel=rbf;, score=0.851 total time=   0.0s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV 1/5] END C=100, class_weight=None, gamma=0.01, kernel=linear;, score=0.866 total time=   0.0s\n",
      "[CV 2/5] END C=100, class_weight=None, gamma=0.01, kernel=linear;, score=0.864 total time=   0.0s\n",
      "[CV 3/5] END C=100, class_weight=None, gamma=0.01, kernel=linear;, score=0.806 total time=   0.0s\n",
      "[CV 4/5] END C=100, class_weight=None, gamma=0.01, kernel=linear;, score=0.871 total time=   0.0s\n",
      "[CV 5/5] END C=100, class_weight=None, gamma=0.01, kernel=linear;, score=0.785 total time=   0.0s\n",
      "[CV 1/5] END C=100, class_weight=None, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=100, class_weight=None, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=100, class_weight=None, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=100, class_weight=None, gamma=0.01, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=100, class_weight=None, gamma=0.01, kernel=poly;, score=0.728 total time=   0.0s\n",
      "[CV 1/5] END C=100, class_weight=None, gamma=0.001, kernel=rbf;, score=0.875 total time=   0.0s\n",
      "[CV 2/5] END C=100, class_weight=None, gamma=0.001, kernel=rbf;, score=0.905 total time=   0.0s\n",
      "[CV 3/5] END C=100, class_weight=None, gamma=0.001, kernel=rbf;, score=0.857 total time=   0.0s\n",
      "[CV 4/5] END C=100, class_weight=None, gamma=0.001, kernel=rbf;, score=0.912 total time=   0.0s\n",
      "[CV 5/5] END C=100, class_weight=None, gamma=0.001, kernel=rbf;, score=0.886 total time=   0.0s\n",
      "[CV 1/5] END C=100, class_weight=None, gamma=0.001, kernel=linear;, score=0.866 total time=   0.0s\n",
      "[CV 2/5] END C=100, class_weight=None, gamma=0.001, kernel=linear;, score=0.864 total time=   0.0s\n",
      "[CV 3/5] END C=100, class_weight=None, gamma=0.001, kernel=linear;, score=0.806 total time=   0.0s\n",
      "[CV 4/5] END C=100, class_weight=None, gamma=0.001, kernel=linear;, score=0.871 total time=   0.0s\n",
      "[CV 5/5] END C=100, class_weight=None, gamma=0.001, kernel=linear;, score=0.785 total time=   0.0s\n",
      "[CV 1/5] END C=100, class_weight=None, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 2/5] END C=100, class_weight=None, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 3/5] END C=100, class_weight=None, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 4/5] END C=100, class_weight=None, gamma=0.001, kernel=poly;, score=0.736 total time=   0.0s\n",
      "[CV 5/5] END C=100, class_weight=None, gamma=0.001, kernel=poly;, score=0.728 total time=   0.0s\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-1 {color: black;background-color: white;}#sk-container-id-1 pre{padding: 0;}#sk-container-id-1 div.sk-toggleable {background-color: white;}#sk-container-id-1 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-1 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-1 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-1 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-1 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-1 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-1 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-1 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-1 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-1 div.sk-item {position: relative;z-index: 1;}#sk-container-id-1 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-1 div.sk-item::before, #sk-container-id-1 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-1 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-1 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-1 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-1 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-1 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-1 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-1 div.sk-label-container {text-align: center;}#sk-container-id-1 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-1 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>GridSearchCV(cv=StratifiedKFold(n_splits=5, random_state=0, shuffle=True),\n",
       "             estimator=SVC(),\n",
       "             param_grid={&#x27;C&#x27;: [0.1, 1, 10, 100],\n",
       "                         &#x27;class_weight&#x27;: [&#x27;balanced&#x27;, None],\n",
       "                         &#x27;gamma&#x27;: [1, 0.1, 0.01, 0.001],\n",
       "                         &#x27;kernel&#x27;: [&#x27;rbf&#x27;, &#x27;linear&#x27;, &#x27;poly&#x27;]},\n",
       "             scoring=&#x27;f1&#x27;, verbose=3)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" ><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">GridSearchCV</label><div class=\"sk-toggleable__content\"><pre>GridSearchCV(cv=StratifiedKFold(n_splits=5, random_state=0, shuffle=True),\n",
       "             estimator=SVC(),\n",
       "             param_grid={&#x27;C&#x27;: [0.1, 1, 10, 100],\n",
       "                         &#x27;class_weight&#x27;: [&#x27;balanced&#x27;, None],\n",
       "                         &#x27;gamma&#x27;: [1, 0.1, 0.01, 0.001],\n",
       "                         &#x27;kernel&#x27;: [&#x27;rbf&#x27;, &#x27;linear&#x27;, &#x27;poly&#x27;]},\n",
       "             scoring=&#x27;f1&#x27;, verbose=3)</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-2\" type=\"checkbox\" ><label for=\"sk-estimator-id-2\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">estimator: SVC</label><div class=\"sk-toggleable__content\"><pre>SVC()</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-3\" type=\"checkbox\" ><label for=\"sk-estimator-id-3\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">SVC</label><div class=\"sk-toggleable__content\"><pre>SVC()</pre></div></div></div></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "GridSearchCV(cv=StratifiedKFold(n_splits=5, random_state=0, shuffle=True),\n",
       "             estimator=SVC(),\n",
       "             param_grid={'C': [0.1, 1, 10, 100],\n",
       "                         'class_weight': ['balanced', None],\n",
       "                         'gamma': [1, 0.1, 0.01, 0.001],\n",
       "                         'kernel': ['rbf', 'linear', 'poly']},\n",
       "             scoring='f1', verbose=3)"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grid = GridSearchCV(SVC(), param_grid, refit = True, scoring = \"f1\", cv=kfold, verbose=3)\n",
    "grid.fit(X_train_resampled, y_train_resampled)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'C': 1, 'class_weight': None, 'gamma': 1, 'kernel': 'poly'}\n"
     ]
    }
   ],
   "source": [
    "print(grid.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.77      0.73      0.75        41\n",
      "           1       0.87      0.89      0.88        83\n",
      "\n",
      "    accuracy                           0.84       124\n",
      "   macro avg       0.82      0.81      0.82       124\n",
      "weighted avg       0.84      0.84      0.84       124\n",
      "\n"
     ]
    }
   ],
   "source": [
    "grid_predictions = grid.predict(X_test_vec)\n",
    "print(classification_report(Y_test, grid_predictions)) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "total number of types extracted is: 7\n",
      "[['fair', 'unlike', 'version', 'loud', 'sound', 'quality', 'poor']]\n",
      "non trovo fair\n",
      "non trovo version\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([0])"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "review = \"\"\"To be fair i don't like this version, it's very loud and the sound quality is poor\"\"\"\n",
    "\n",
    "review_tok = tokenize_list_of_text([review], custom_stopwords, False, 2)[1]\n",
    "print(review_tok)\n",
    "review_vec = review_vectors(review_tok[0], w2v_model.vector_size, [1 for x in range(len(review_tok[0]))])\n",
    "pred = grid.predict(review_vec)\n",
    "pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.7"
  },
  "vscode": {
   "interpreter": {
    "hash": "43973268aa23c7cf4b4cafe0e819fac9b2412fcc991a21988492c3271a4e3f9b"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}