naive_bayes.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_selected = pd.read_csv('rusentiment_preselected_posts.csv')\n",
    "df_random = pd.read_csv('rusentiment_random_posts.csv')\n",
    "df_test = pd.read_csv('rusentiment_test.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train = pd.concat((df_selected, df_random))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Original code https://github.com/akutuzov/webvectors/blob/master/preprocessing/rus_preprocessing_udpipe.py\n",
    "from ufal.udpipe import Model, Pipeline\n",
    "\n",
    "#! wget http://rusvectores.org/static/models/udpipe_syntagrus.model\n",
    "udpipe_filename = 'udpipe_syntagrus.model'\n",
    "model = Model.load(udpipe_filename)\n",
    "process_pipeline = Pipeline(model, 'tokenize', Pipeline.DEFAULT, Pipeline.DEFAULT, 'conllu')\n",
    "\n",
    "def tag_ud(pipeline, text='Text in string format', pos=True):\n",
    "    # если частеречные тэги не нужны (например, их нет в модели), выставьте pos=False\n",
    "    # в этом случае на выход будут поданы только леммы\n",
    "\n",
    "    # обрабатываем текст, получаем результат в формате conllu:\n",
    "    processed = pipeline.process(text)\n",
    "\n",
    "    # пропускаем строки со служебной информацией:\n",
    "    content = [l for l in processed.split('\\n') if not l.startswith('#')]\n",
    "\n",
    "    # извлекаем из обработанного текста лемму и тэг\n",
    "    tagged = [w.split('\\t')[2].lower() + '_' + w.split('\\t')[3] for w in content if w]\n",
    "    result = []\n",
    "    for t in tagged:\n",
    "        if (len(t.split('_')) == 2):\n",
    "            lemma, tag = t.split('_')\n",
    "            # убираем числительные и все знаки пунктуации кроме скобок\n",
    "            if tag == 'NUM' or (tag == 'PUNCT' and lemma not in ['(', ')']):\n",
    "                continue\n",
    "            if pos:\n",
    "                result.append(lemma + '_' + tag)\n",
    "            else:\n",
    "                result.append(lemma)\n",
    "    return result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train['lemmatized_tokens'] = df_train.apply(lambda x: tag_ud(process_pipeline, x['text'], pos=False), axis=1)\n",
    "df_test['lemmatized_tokens'] = df_test.apply(lambda x: tag_ud(process_pipeline, x['text'], pos=False), axis=1)\n",
    "\n",
    "df_train['lemmatized_text'] = df_train['lemmatized_tokens'].str.join(' ')\n",
    "df_test['lemmatized_text'] = df_test['lemmatized_tokens'].str.join(' ')\n",
    "\n",
    "df_train = df_train[~df_train.lemmatized_text.isnull()]\n",
    "df_test = df_test[~df_test.lemmatized_text.isnull()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import preprocessing\n",
    "le = preprocessing.LabelEncoder()\n",
    "\n",
    "X_train = df_train['lemmatized_text'].values\n",
    "X_test = df_test['lemmatized_text'].values\n",
    "\n",
    "y_train  = le.fit_transform(df_train['label'].values)\n",
    "y_test  = le.transform(df_test['label'].values) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 58min 57s, sys: 31.5 s, total: 59min 29s\n",
      "Wall time: 51min 33s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "from sklearn.pipeline import make_pipeline\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.naive_bayes import MultinomialNB\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.model_selection import StratifiedKFold\n",
    "from scipy.stats import expon\n",
    "\n",
    "\n",
    "exp = expon(scale=2)\n",
    "NB_pipeline = Pipeline([('tfidf', TfidfVectorizer()), ('multiNB', MultinomialNB(class_prior=None))])\n",
    "param_grid = {'tfidf__ngram_range':[(1,2), (1,1)],\n",
    "              'tfidf__sublinear_tf':[True, False],\n",
    "              'tfidf__min_df':np.arange(2,5,1),\n",
    "              'multiNB__alpha': exp.rvs(size=10), \n",
    "              'multiNB__fit_prior':[True, False]\n",
    "              }\n",
    "\n",
    "skf = StratifiedKFold(n_splits=5, random_state=42)\n",
    "gr = GridSearchCV(NB_pipeline, param_grid, cv=skf, verbose=0)\n",
    "gr_model = gr.fit(X_train, y_train)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "result = gr_model.cv_results_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "rows = list ()\n",
    "for model in result['params']:\n",
    "    row = []\n",
    "    for key, val in model.items():\n",
    "        if key=='classifier':\n",
    "            v_str = (val.__str__())\n",
    "            row.append(v_str [ : v_str.find('(')])\n",
    "        else:\n",
    "            param = key.__str__()\n",
    "            param = param [ param.find('__')+2 :]\n",
    "            row.append(param + ' = ' + val.__str__())\n",
    "    rows.append(row)\n",
    "\n",
    "df = pd.DataFrame(rows)\n",
    "df['score'] = result['mean_test_score']\n",
    "df['time'] = result['mean_fit_time']\n",
    "\n",
    "df = df.sort_values(by=['score'], ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>score</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>173</th>\n",
       "      <td>alpha = 0.21970331388934286</td>\n",
       "      <td>fit_prior = True</td>\n",
       "      <td>min_df = 3</td>\n",
       "      <td>ngram_range = (1, 2)</td>\n",
       "      <td>sublinear_tf = False</td>\n",
       "      <td>0.582182</td>\n",
       "      <td>2.079544</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>169</th>\n",
       "      <td>alpha = 0.21970331388934286</td>\n",
       "      <td>fit_prior = True</td>\n",
       "      <td>min_df = 2</td>\n",
       "      <td>ngram_range = (1, 2)</td>\n",
       "      <td>sublinear_tf = False</td>\n",
       "      <td>0.581437</td>\n",
       "      <td>1.933962</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>168</th>\n",
       "      <td>alpha = 0.21970331388934286</td>\n",
       "      <td>fit_prior = True</td>\n",
       "      <td>min_df = 2</td>\n",
       "      <td>ngram_range = (1, 2)</td>\n",
       "      <td>sublinear_tf = True</td>\n",
       "      <td>0.581048</td>\n",
       "      <td>2.009645</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>177</th>\n",
       "      <td>alpha = 0.21970331388934286</td>\n",
       "      <td>fit_prior = True</td>\n",
       "      <td>min_df = 4</td>\n",
       "      <td>ngram_range = (1, 2)</td>\n",
       "      <td>sublinear_tf = False</td>\n",
       "      <td>0.580870</td>\n",
       "      <td>2.298541</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>176</th>\n",
       "      <td>alpha = 0.21970331388934286</td>\n",
       "      <td>fit_prior = True</td>\n",
       "      <td>min_df = 4</td>\n",
       "      <td>ngram_range = (1, 2)</td>\n",
       "      <td>sublinear_tf = True</td>\n",
       "      <td>0.580729</td>\n",
       "      <td>2.034982</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>alpha = 0.12162256502136387</td>\n",
       "      <td>fit_prior = True</td>\n",
       "      <td>min_df = 2</td>\n",
       "      <td>ngram_range = (1, 2)</td>\n",
       "      <td>sublinear_tf = True</td>\n",
       "      <td>0.580091</td>\n",
       "      <td>1.889471</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>alpha = 0.12162256502136387</td>\n",
       "      <td>fit_prior = True</td>\n",
       "      <td>min_df = 2</td>\n",
       "      <td>ngram_range = (1, 2)</td>\n",
       "      <td>sublinear_tf = False</td>\n",
       "      <td>0.579843</td>\n",
       "      <td>1.916305</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>alpha = 0.12162256502136387</td>\n",
       "      <td>fit_prior = True</td>\n",
       "      <td>min_df = 3</td>\n",
       "      <td>ngram_range = (1, 2)</td>\n",
       "      <td>sublinear_tf = False</td>\n",
       "      <td>0.579382</td>\n",
       "      <td>1.926445</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>alpha = 0.12162256502136387</td>\n",
       "      <td>fit_prior = True</td>\n",
       "      <td>min_df = 3</td>\n",
       "      <td>ngram_range = (1, 2)</td>\n",
       "      <td>sublinear_tf = True</td>\n",
       "      <td>0.578957</td>\n",
       "      <td>1.845136</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>200</th>\n",
       "      <td>alpha = 0.562308750517385</td>\n",
       "      <td>fit_prior = True</td>\n",
       "      <td>min_df = 4</td>\n",
       "      <td>ngram_range = (1, 2)</td>\n",
       "      <td>sublinear_tf = True</td>\n",
       "      <td>0.578496</td>\n",
       "      <td>2.015164</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>57</th>\n",
       "      <td>alpha = 0.12162256502136387</td>\n",
       "      <td>fit_prior = True</td>\n",
       "      <td>min_df = 4</td>\n",
       "      <td>ngram_range = (1, 2)</td>\n",
       "      <td>sublinear_tf = False</td>\n",
       "      <td>0.577894</td>\n",
       "      <td>1.910072</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>201</th>\n",
       "      <td>alpha = 0.562308750517385</td>\n",
       "      <td>fit_prior = True</td>\n",
       "      <td>min_df = 4</td>\n",
       "      <td>ngram_range = (1, 2)</td>\n",
       "      <td>sublinear_tf = False</td>\n",
       "      <td>0.577681</td>\n",
       "      <td>2.041291</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>alpha = 0.12162256502136387</td>\n",
       "      <td>fit_prior = True</td>\n",
       "      <td>min_df = 4</td>\n",
       "      <td>ngram_range = (1, 2)</td>\n",
       "      <td>sublinear_tf = True</td>\n",
       "      <td>0.577645</td>\n",
       "      <td>1.886206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>175</th>\n",
       "      <td>alpha = 0.21970331388934286</td>\n",
       "      <td>fit_prior = True</td>\n",
       "      <td>min_df = 3</td>\n",
       "      <td>ngram_range = (1, 1)</td>\n",
       "      <td>sublinear_tf = False</td>\n",
       "      <td>0.576689</td>\n",
       "      <td>0.891954</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>174</th>\n",
       "      <td>alpha = 0.21970331388934286</td>\n",
       "      <td>fit_prior = True</td>\n",
       "      <td>min_df = 3</td>\n",
       "      <td>ngram_range = (1, 1)</td>\n",
       "      <td>sublinear_tf = True</td>\n",
       "      <td>0.576263</td>\n",
       "      <td>0.838102</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>196</th>\n",
       "      <td>alpha = 0.562308750517385</td>\n",
       "      <td>fit_prior = True</td>\n",
       "      <td>min_df = 3</td>\n",
       "      <td>ngram_range = (1, 2)</td>\n",
       "      <td>sublinear_tf = True</td>\n",
       "      <td>0.576051</td>\n",
       "      <td>2.250868</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>197</th>\n",
       "      <td>alpha = 0.562308750517385</td>\n",
       "      <td>fit_prior = True</td>\n",
       "      <td>min_df = 3</td>\n",
       "      <td>ngram_range = (1, 2)</td>\n",
       "      <td>sublinear_tf = False</td>\n",
       "      <td>0.575661</td>\n",
       "      <td>1.928009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>178</th>\n",
       "      <td>alpha = 0.21970331388934286</td>\n",
       "      <td>fit_prior = True</td>\n",
       "      <td>min_df = 4</td>\n",
       "      <td>ngram_range = (1, 1)</td>\n",
       "      <td>sublinear_tf = True</td>\n",
       "      <td>0.575023</td>\n",
       "      <td>0.720727</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>54</th>\n",
       "      <td>alpha = 0.12162256502136387</td>\n",
       "      <td>fit_prior = True</td>\n",
       "      <td>min_df = 3</td>\n",
       "      <td>ngram_range = (1, 1)</td>\n",
       "      <td>sublinear_tf = True</td>\n",
       "      <td>0.574952</td>\n",
       "      <td>0.703913</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                               0                 1           2  \\\n",
       "173  alpha = 0.21970331388934286  fit_prior = True  min_df = 3   \n",
       "169  alpha = 0.21970331388934286  fit_prior = True  min_df = 2   \n",
       "168  alpha = 0.21970331388934286  fit_prior = True  min_df = 2   \n",
       "177  alpha = 0.21970331388934286  fit_prior = True  min_df = 4   \n",
       "176  alpha = 0.21970331388934286  fit_prior = True  min_df = 4   \n",
       "48   alpha = 0.12162256502136387  fit_prior = True  min_df = 2   \n",
       "49   alpha = 0.12162256502136387  fit_prior = True  min_df = 2   \n",
       "53   alpha = 0.12162256502136387  fit_prior = True  min_df = 3   \n",
       "52   alpha = 0.12162256502136387  fit_prior = True  min_df = 3   \n",
       "200    alpha = 0.562308750517385  fit_prior = True  min_df = 4   \n",
       "57   alpha = 0.12162256502136387  fit_prior = True  min_df = 4   \n",
       "201    alpha = 0.562308750517385  fit_prior = True  min_df = 4   \n",
       "56   alpha = 0.12162256502136387  fit_prior = True  min_df = 4   \n",
       "175  alpha = 0.21970331388934286  fit_prior = True  min_df = 3   \n",
       "174  alpha = 0.21970331388934286  fit_prior = True  min_df = 3   \n",
       "196    alpha = 0.562308750517385  fit_prior = True  min_df = 3   \n",
       "197    alpha = 0.562308750517385  fit_prior = True  min_df = 3   \n",
       "178  alpha = 0.21970331388934286  fit_prior = True  min_df = 4   \n",
       "54   alpha = 0.12162256502136387  fit_prior = True  min_df = 3   \n",
       "\n",
       "                        3                     4     score      time  \n",
       "173  ngram_range = (1, 2)  sublinear_tf = False  0.582182  2.079544  \n",
       "169  ngram_range = (1, 2)  sublinear_tf = False  0.581437  1.933962  \n",
       "168  ngram_range = (1, 2)   sublinear_tf = True  0.581048  2.009645  \n",
       "177  ngram_range = (1, 2)  sublinear_tf = False  0.580870  2.298541  \n",
       "176  ngram_range = (1, 2)   sublinear_tf = True  0.580729  2.034982  \n",
       "48   ngram_range = (1, 2)   sublinear_tf = True  0.580091  1.889471  \n",
       "49   ngram_range = (1, 2)  sublinear_tf = False  0.579843  1.916305  \n",
       "53   ngram_range = (1, 2)  sublinear_tf = False  0.579382  1.926445  \n",
       "52   ngram_range = (1, 2)   sublinear_tf = True  0.578957  1.845136  \n",
       "200  ngram_range = (1, 2)   sublinear_tf = True  0.578496  2.015164  \n",
       "57   ngram_range = (1, 2)  sublinear_tf = False  0.577894  1.910072  \n",
       "201  ngram_range = (1, 2)  sublinear_tf = False  0.577681  2.041291  \n",
       "56   ngram_range = (1, 2)   sublinear_tf = True  0.577645  1.886206  \n",
       "175  ngram_range = (1, 1)  sublinear_tf = False  0.576689  0.891954  \n",
       "174  ngram_range = (1, 1)   sublinear_tf = True  0.576263  0.838102  \n",
       "196  ngram_range = (1, 2)   sublinear_tf = True  0.576051  2.250868  \n",
       "197  ngram_range = (1, 2)  sublinear_tf = False  0.575661  1.928009  \n",
       "178  ngram_range = (1, 1)   sublinear_tf = True  0.575023  0.720727  \n",
       "54   ngram_range = (1, 1)   sublinear_tf = True  0.574952  0.703913  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[1:20]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "fit_prior = True, равно как и ngram_range=(1,2) дают стабильный прирост точности. Минимальное количество документов варьируется от 2 до 4. В то же время варьирование alpha и sublinear_tf не дают значительного прироста качества предсказывания."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "({'multiNB__alpha': 0.21970331388934286,\n",
       "  'multiNB__fit_prior': True,\n",
       "  'tfidf__min_df': 3,\n",
       "  'tfidf__ngram_range': (1, 2),\n",
       "  'tfidf__sublinear_tf': True},\n",
       " 0.5829966687929691)"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gr_model.best_params_, gr_model.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "best_pipe = Pipeline([('tfidf',TfidfVectorizer(ngram_range=(1,2), min_df=3, sublinear_tf=True)),\n",
    "                              ('multiNB', MultinomialNB(alpha=0.21970331388934286, fit_prior = True))])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Precision: 0.6408\n",
      "Recall: 0.6593\n",
      "f1-score: 0.6331\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import f1_score, precision_score, recall_score, confusion_matrix\n",
    "\n",
    "best_pipe.fit(X_train, y_train)\n",
    "y_pred = best_pipe.predict(X_test)\n",
    "def score_model(y_pred, y_true):\n",
    "    average = 'weighted'\n",
    "    f1 = f1_score(y_true, y_pred, average=average)\n",
    "    precision = precision_score(y_true, y_pred, average=average)\n",
    "    recall = recall_score(y_true, y_pred, average=average)\n",
    "    return f1, precision, recall\n",
    "\n",
    "f1, precision, recall = score_model(y_pred,y_test)\n",
    "print(f'Precision: {precision:.4f}')\n",
    "print(f'Recall: {recall:.4f}')\n",
    "print(f'f1-score: {f1:.4f}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.metrics import confusion_matrix\n",
    "conf_mat = confusion_matrix(y_test, y_pred)\n",
    "conf_mat_ratios = np.round(conf_mat / np.sum(conf_mat, axis=0),3)\n",
    "fig, ax = plt.subplots(figsize=(10,8))\n",
    "\n",
    "labels = sorted(list(set(le.inverse_transform(y_pred))))\n",
    "sns.heatmap(conf_mat_ratios, annot=True, center=0, xticklabels=labels, yticklabels=labels)\n",
    "plt.ylabel('Actual')\n",
    "plt.xlabel('Predicted')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred_labels = le.inverse_transform(y_pred)\n",
    "y_test_labels = le.inverse_transform(y_test)\n",
    "\n",
    "results = pd.DataFrame({'text':df_test.text, 'lemmatized_tokens':df_test.lemmatized_tokens,'predicted':y_pred_labels, 'actual':y_test_labels})\n",
    "errors = results[results.predicted != results.actual]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(12, 8))\n",
    "freq = pd.DataFrame({'Predicted': errors.predicted.value_counts(), 'Actual':errors.actual.value_counts()})\n",
    "freq.sort_values(by='Predicted', ascending=False).plot.bar(rot=0, ax=ax);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "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>text</th>\n",
       "      <th>lemmatized_tokens</th>\n",
       "      <th>predicted</th>\n",
       "      <th>actual</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Мы просто с  [id12210212|Елисеем]  психанули )...</td>\n",
       "      <td>[мы, просто, с, id12210212|елисей, психануть, ...</td>\n",
       "      <td>positive</td>\n",
       "      <td>skip</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Наши красавцы. 1:3 в первом периоде и 7:3 в ит...</td>\n",
       "      <td>[наш, красавец, в, первый, период, и, в, итог]</td>\n",
       "      <td>neutral</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>нельзя общаться с людьми, которые нуждаются в ...</td>\n",
       "      <td>[нельзя, общаться, с, человек, который, нуждат...</td>\n",
       "      <td>neutral</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>Я не писал такое оскобление всем!!!!\\nМеня, ви...</td>\n",
       "      <td>[я, не, писать, такой, оскоблений, все, я, вид...</td>\n",
       "      <td>neutral</td>\n",
       "      <td>skip</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>...и так начало мини отпуска))И новых приключе...</td>\n",
       "      <td>[...ь, так, начинать, минять, отпуск, ), ), и,...</td>\n",
       "      <td>neutral</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>Кинул в чс обиделся холоп понял что недостоин ...</td>\n",
       "      <td>[кидать, в, чс, обижаться, холоп, понимать, чт...</td>\n",
       "      <td>neutral</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>Лучше бы у Тани не было никакого оргазма: услы...</td>\n",
       "      <td>[хороший, бы, в, таня, не, быть, никакой, орга...</td>\n",
       "      <td>neutral</td>\n",
       "      <td>skip</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>по кругу, твою подругу КСЮХУ....))))))ахахахаахах</td>\n",
       "      <td>[по, круг, твой, подруга, ксюх, ), ), ), ), ),...</td>\n",
       "      <td>positive</td>\n",
       "      <td>skip</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>без тебя_(</td>\n",
       "      <td>[без, (]</td>\n",
       "      <td>neutral</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>я прыгнул в бездну из уюта пора готовиться к з...</td>\n",
       "      <td>[я, прыгнуть, в, бездна, из, уют, пора, готови...</td>\n",
       "      <td>neutral</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>амммм, Каринка)))</td>\n",
       "      <td>[амммма, каринк, ), ), )]</td>\n",
       "      <td>neutral</td>\n",
       "      <td>skip</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>З 19 вересня тебе!</td>\n",
       "      <td>[с, вересня, ты]</td>\n",
       "      <td>neutral</td>\n",
       "      <td>skip</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>Мы растем!</td>\n",
       "      <td>[мы, растать]</td>\n",
       "      <td>neutral</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>на аву прямо сейчааас!</td>\n",
       "      <td>[на, ава, прямо, сейчааасать]</td>\n",
       "      <td>positive</td>\n",
       "      <td>neutral</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>хоть и качество плохое, но обещал!!!!) ребятки...</td>\n",
       "      <td>[хоть, и, качество, плохой, но, обещать, ), ре...</td>\n",
       "      <td>positive</td>\n",
       "      <td>neutral</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>нахуй Лешего на аву поставила? О_о</td>\n",
       "      <td>[нахуй, леший, на, ава, ставить]</td>\n",
       "      <td>neutral</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>Я доверяю немногим. Но эти немногие, мне реаль...</td>\n",
       "      <td>[я, доверять, немногое, но, этот, немногие, я,...</td>\n",
       "      <td>neutral</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>Димон, в Дубаях еть крутой шейх, который на те...</td>\n",
       "      <td>[димон, в, дубаи, еть, крутой, шейх, который, ...</td>\n",
       "      <td>neutral</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>ПИДОР го скайп</td>\n",
       "      <td>[пидор, 1-й, скайп]</td>\n",
       "      <td>negative</td>\n",
       "      <td>skip</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>чувак тупо лол___)) ахаххахаах</td>\n",
       "      <td>[чувак, тупо, ), ), ахаххаха]</td>\n",
       "      <td>neutral</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>Вот и моя первая выставка!</td>\n",
       "      <td>[вот, и, мой, первый, выставка]</td>\n",
       "      <td>neutral</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>Я жаловался на мусор в ростовских рощах и парк...</td>\n",
       "      <td>[я, жаловаться, на, мусор, в, ростовский, роща...</td>\n",
       "      <td>neutral</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>54</th>\n",
       "      <td>Кто-то заходит на мою страничку, с умилением с...</td>\n",
       "      <td>[кто-то, заходить, на, мой, страничка, с, умил...</td>\n",
       "      <td>neutral</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>57</th>\n",
       "      <td>Привет красавица ;)\\nЛиночка надеюсь я не ошиб...</td>\n",
       "      <td>[привет, красавица, ), линочка, надеяться, я, ...</td>\n",
       "      <td>positive</td>\n",
       "      <td>speech</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>63</th>\n",
       "      <td>После 6-ти часовой задержки в Бангкоке и второ...</td>\n",
       "      <td>[после, 6, часовой, задержка, в, бангкок, и, в...</td>\n",
       "      <td>neutral</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>67</th>\n",
       "      <td>Друзья - это редкие драгоценности, они принося...</td>\n",
       "      <td>[друг, это, редкий, драгоценность, они, принос...</td>\n",
       "      <td>neutral</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>Заболiв!(((</td>\n",
       "      <td>[заболiв, (, (, (]</td>\n",
       "      <td>neutral</td>\n",
       "      <td>skip</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70</th>\n",
       "      <td>- Мадам, я вчера слышал ваше пение...\\n- О, чт...</td>\n",
       "      <td>[мадам, я, вчера, слышать, ваш, пение, о, что,...</td>\n",
       "      <td>negative</td>\n",
       "      <td>skip</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>71</th>\n",
       "      <td>Всех, кто предал – прощаю!\\nВсем, кто ушел – ж...</td>\n",
       "      <td>[все, кто, предавать, прощать, все, кто, уходи...</td>\n",
       "      <td>neutral</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>72</th>\n",
       "      <td>Ужассс извращёная девочка!</td>\n",
       "      <td>[ужассс, извращёный, девочка]</td>\n",
       "      <td>neutral</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2893</th>\n",
       "      <td>любимому две ложечки  сахара!!!</td>\n",
       "      <td>[любить, ложечка, сахар]</td>\n",
       "      <td>neutral</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2903</th>\n",
       "      <td>приспичило белке купить под конец осени кепку,...</td>\n",
       "      <td>[приспичить, белок, покупать, под, конец, осен...</td>\n",
       "      <td>neutral</td>\n",
       "      <td>skip</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2904</th>\n",
       "      <td>Вместо лиц - аватары..вместо настроения - стат...</td>\n",
       "      <td>[вместо, лицо, аватары., .вместо, настроение, ...</td>\n",
       "      <td>neutral</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2905</th>\n",
       "      <td>Вятка! Спасибо за великолепные выходные! Это б...</td>\n",
       "      <td>[вятка, спасибо, за, великолепный, выходной, э...</td>\n",
       "      <td>positive</td>\n",
       "      <td>speech</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2907</th>\n",
       "      <td>Поздравте...у меня сегодня племянница радилась:)</td>\n",
       "      <td>[поздравте., ., я, сегодня, племянница, радить...</td>\n",
       "      <td>neutral</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2908</th>\n",
       "      <td>Я люблю свою сестричку!!!!\\nВона в мене найкра...</td>\n",
       "      <td>[я, любить, свой, сестричка, вона, в, я, цьом**]</td>\n",
       "      <td>positive</td>\n",
       "      <td>skip</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2909</th>\n",
       "      <td>Душа хочет романтики...попа требует приключени...</td>\n",
       "      <td>[душа, хотеть, романтики., .поп, требовать, пр...</td>\n",
       "      <td>negative</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2911</th>\n",
       "      <td>Вот оно - искусство!!! ))</td>\n",
       "      <td>[вот, оно, искусство, ), )]</td>\n",
       "      <td>neutral</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2916</th>\n",
       "      <td>З Святом вас Святого Миколая!!!!Най принесе ва...</td>\n",
       "      <td>[с, святой, вы, святой, миколай, най, принес, ...</td>\n",
       "      <td>speech</td>\n",
       "      <td>skip</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2917</th>\n",
       "      <td>помню как было тяжело , но очень хочу это все ...</td>\n",
       "      <td>[помнить, как, быть, тяжелый, но, очень, хотет...</td>\n",
       "      <td>positive</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2920</th>\n",
       "      <td>Когда-нибудь и мы поженимся.</td>\n",
       "      <td>[когда-нибудь, и, мы, пожениться]</td>\n",
       "      <td>positive</td>\n",
       "      <td>skip</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2921</th>\n",
       "      <td>Я прощаю одиночество и грусть, ты сказал, что ...</td>\n",
       "      <td>[я, прощать, одиночество, и, грусть, ты, сказа...</td>\n",
       "      <td>positive</td>\n",
       "      <td>neutral</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2925</th>\n",
       "      <td>Иногда увольняться полезно! ;)))</td>\n",
       "      <td>[иногда, увольнять, полезный, ), ), )]</td>\n",
       "      <td>neutral</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2931</th>\n",
       "      <td>НУФВСё!!!\\n\\nГАМОВЕР УЧЕБЕ \\nя теперь маджистер</td>\n",
       "      <td>[нуфвсё, гамовирать, учеба, я, теперь, маджист...</td>\n",
       "      <td>neutral</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2932</th>\n",
       "      <td>Сегодня прогулялся с довольно интересной девуш...</td>\n",
       "      <td>[сегодня, прогуляться, с, довольно, интересный...</td>\n",
       "      <td>neutral</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2934</th>\n",
       "      <td>ты чоооо-о не поздравил меня с этим великим пр...</td>\n",
       "      <td>[ты, чоооо-о, не, поздравлять, я, с, это, вели...</td>\n",
       "      <td>speech</td>\n",
       "      <td>neutral</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2937</th>\n",
       "      <td>многие пиздят в инете за этот фильм) от создат...</td>\n",
       "      <td>[многие, пиздить, в, инет, за, этот, фильм, ),...</td>\n",
       "      <td>neutral</td>\n",
       "      <td>skip</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2940</th>\n",
       "      <td>Танюшечка❤</td>\n",
       "      <td>[танюшечка❤]</td>\n",
       "      <td>neutral</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2942</th>\n",
       "      <td>кшод4 аркгаркгаи</td>\n",
       "      <td>[аркгаркгай]</td>\n",
       "      <td>neutral</td>\n",
       "      <td>skip</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2944</th>\n",
       "      <td>Завтра в мене перший екзамен!</td>\n",
       "      <td>[завтра, в, я, перать, екзамен]</td>\n",
       "      <td>neutral</td>\n",
       "      <td>skip</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2946</th>\n",
       "      <td>стою смотрю в бесконечность.. задумался о смыс...</td>\n",
       "      <td>[стоить, смотреть, в, бесконечность, задумыват...</td>\n",
       "      <td>positive</td>\n",
       "      <td>neutral</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2949</th>\n",
       "      <td>Тащуся 8-)</td>\n",
       "      <td>[тащуся, )]</td>\n",
       "      <td>neutral</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2951</th>\n",
       "      <td>ПАЗДРАВЛЯЮ!!!</td>\n",
       "      <td>[паздравляю]</td>\n",
       "      <td>neutral</td>\n",
       "      <td>speech</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2954</th>\n",
       "      <td>ВАЩЕЕЕ КРАСАВАА!!!</td>\n",
       "      <td>[ващео, красаваа.]</td>\n",
       "      <td>neutral</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2955</th>\n",
       "      <td>В эту песню я просто влюблен еще с 90-х. Мы го...</td>\n",
       "      <td>[в, этот, песня, я, просто, влюбить, еще, с, 9...</td>\n",
       "      <td>neutral</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2957</th>\n",
       "      <td>אימון קרב מגע מתקדם ביחידה מובחרת - Krav Maga</td>\n",
       "      <td>[אימון, קרב, מגע, מתקדם, ביחידה, מובחרת, krav,...</td>\n",
       "      <td>neutral</td>\n",
       "      <td>skip</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2959</th>\n",
       "      <td>Бля я и забыл, какой он ахуенный!</td>\n",
       "      <td>[бля, я, и, забывать, какой, он, ахуать]</td>\n",
       "      <td>negative</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2961</th>\n",
       "      <td>Ярких эмоций, незабываемых встреч и полета душ...</td>\n",
       "      <td>[яркий, эмоция, незабываемый, встреча, и, поле...</td>\n",
       "      <td>neutral</td>\n",
       "      <td>speech</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2964</th>\n",
       "      <td>Ты одна половинка - я вторая половинка...... А...</td>\n",
       "      <td>[ты, один, половинка, я, второй, половинка, а,...</td>\n",
       "      <td>positive</td>\n",
       "      <td>skip</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2966</th>\n",
       "      <td>Ахахахахах.Прикольно попереписувались.\\nаааааа...</td>\n",
       "      <td>[ахахахахы, прикольно, попереписуваться]</td>\n",
       "      <td>positive</td>\n",
       "      <td>skip</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1011 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                   text  \\\n",
       "3     Мы просто с  [id12210212|Елисеем]  психанули )...   \n",
       "5     Наши красавцы. 1:3 в первом периоде и 7:3 в ит...   \n",
       "8     нельзя общаться с людьми, которые нуждаются в ...   \n",
       "17    Я не писал такое оскобление всем!!!!\\nМеня, ви...   \n",
       "18    ...и так начало мини отпуска))И новых приключе...   \n",
       "19    Кинул в чс обиделся холоп понял что недостоин ...   \n",
       "21    Лучше бы у Тани не было никакого оргазма: услы...   \n",
       "25    по кругу, твою подругу КСЮХУ....))))))ахахахаахах   \n",
       "26                                           без тебя_(   \n",
       "27    я прыгнул в бездну из уюта пора готовиться к з...   \n",
       "29                                    амммм, Каринка)))   \n",
       "30                                   З 19 вересня тебе!   \n",
       "31                                           Мы растем!   \n",
       "33                               на аву прямо сейчааас!   \n",
       "38    хоть и качество плохое, но обещал!!!!) ребятки...   \n",
       "40                   нахуй Лешего на аву поставила? О_о   \n",
       "43    Я доверяю немногим. Но эти немногие, мне реаль...   \n",
       "46    Димон, в Дубаях еть крутой шейх, который на те...   \n",
       "47                                       ПИДОР го скайп   \n",
       "48                       чувак тупо лол___)) ахаххахаах   \n",
       "51                           Вот и моя первая выставка!   \n",
       "52    Я жаловался на мусор в ростовских рощах и парк...   \n",
       "54    Кто-то заходит на мою страничку, с умилением с...   \n",
       "57    Привет красавица ;)\\nЛиночка надеюсь я не ошиб...   \n",
       "63    После 6-ти часовой задержки в Бангкоке и второ...   \n",
       "67    Друзья - это редкие драгоценности, они принося...   \n",
       "68                                          Заболiв!(((   \n",
       "70    - Мадам, я вчера слышал ваше пение...\\n- О, чт...   \n",
       "71    Всех, кто предал – прощаю!\\nВсем, кто ушел – ж...   \n",
       "72                           Ужассс извращёная девочка!   \n",
       "...                                                 ...   \n",
       "2893                    любимому две ложечки  сахара!!!   \n",
       "2903  приспичило белке купить под конец осени кепку,...   \n",
       "2904  Вместо лиц - аватары..вместо настроения - стат...   \n",
       "2905  Вятка! Спасибо за великолепные выходные! Это б...   \n",
       "2907   Поздравте...у меня сегодня племянница радилась:)   \n",
       "2908  Я люблю свою сестричку!!!!\\nВона в мене найкра...   \n",
       "2909  Душа хочет романтики...попа требует приключени...   \n",
       "2911                          Вот оно - искусство!!! ))   \n",
       "2916  З Святом вас Святого Миколая!!!!Най принесе ва...   \n",
       "2917  помню как было тяжело , но очень хочу это все ...   \n",
       "2920                       Когда-нибудь и мы поженимся.   \n",
       "2921  Я прощаю одиночество и грусть, ты сказал, что ...   \n",
       "2925                   Иногда увольняться полезно! ;)))   \n",
       "2931    НУФВСё!!!\\n\\nГАМОВЕР УЧЕБЕ \\nя теперь маджистер   \n",
       "2932  Сегодня прогулялся с довольно интересной девуш...   \n",
       "2934  ты чоооо-о не поздравил меня с этим великим пр...   \n",
       "2937  многие пиздят в инете за этот фильм) от создат...   \n",
       "2940                                         Танюшечка❤   \n",
       "2942                                   кшод4 аркгаркгаи   \n",
       "2944                      Завтра в мене перший екзамен!   \n",
       "2946  стою смотрю в бесконечность.. задумался о смыс...   \n",
       "2949                                         Тащуся 8-)   \n",
       "2951                                      ПАЗДРАВЛЯЮ!!!   \n",
       "2954                                 ВАЩЕЕЕ КРАСАВАА!!!   \n",
       "2955  В эту песню я просто влюблен еще с 90-х. Мы го...   \n",
       "2957      אימון קרב מגע מתקדם ביחידה מובחרת - Krav Maga   \n",
       "2959                  Бля я и забыл, какой он ахуенный!   \n",
       "2961  Ярких эмоций, незабываемых встреч и полета душ...   \n",
       "2964  Ты одна половинка - я вторая половинка...... А...   \n",
       "2966  Ахахахахах.Прикольно попереписувались.\\nаааааа...   \n",
       "\n",
       "                                      lemmatized_tokens predicted    actual  \n",
       "3     [мы, просто, с, id12210212|елисей, психануть, ...  positive      skip  \n",
       "5        [наш, красавец, в, первый, период, и, в, итог]   neutral  positive  \n",
       "8     [нельзя, общаться, с, человек, который, нуждат...   neutral  negative  \n",
       "17    [я, не, писать, такой, оскоблений, все, я, вид...   neutral      skip  \n",
       "18    [...ь, так, начинать, минять, отпуск, ), ), и,...   neutral  positive  \n",
       "19    [кидать, в, чс, обижаться, холоп, понимать, чт...   neutral  negative  \n",
       "21    [хороший, бы, в, таня, не, быть, никакой, орга...   neutral      skip  \n",
       "25    [по, круг, твой, подруга, ксюх, ), ), ), ), ),...  positive      skip  \n",
       "26                                             [без, (]   neutral  negative  \n",
       "27    [я, прыгнуть, в, бездна, из, уют, пора, готови...   neutral  negative  \n",
       "29                            [амммма, каринк, ), ), )]   neutral      skip  \n",
       "30                                     [с, вересня, ты]   neutral      skip  \n",
       "31                                        [мы, растать]   neutral  positive  \n",
       "33                        [на, ава, прямо, сейчааасать]  positive   neutral  \n",
       "38    [хоть, и, качество, плохой, но, обещать, ), ре...  positive   neutral  \n",
       "40                     [нахуй, леший, на, ава, ставить]   neutral  negative  \n",
       "43    [я, доверять, немногое, но, этот, немногие, я,...   neutral  positive  \n",
       "46    [димон, в, дубаи, еть, крутой, шейх, который, ...   neutral  positive  \n",
       "47                                  [пидор, 1-й, скайп]  negative      skip  \n",
       "48                        [чувак, тупо, ), ), ахаххаха]   neutral  positive  \n",
       "51                      [вот, и, мой, первый, выставка]   neutral  positive  \n",
       "52    [я, жаловаться, на, мусор, в, ростовский, роща...   neutral  negative  \n",
       "54    [кто-то, заходить, на, мой, страничка, с, умил...   neutral  positive  \n",
       "57    [привет, красавица, ), линочка, надеяться, я, ...  positive    speech  \n",
       "63    [после, 6, часовой, задержка, в, бангкок, и, в...   neutral  positive  \n",
       "67    [друг, это, редкий, драгоценность, они, принос...   neutral  positive  \n",
       "68                                   [заболiв, (, (, (]   neutral      skip  \n",
       "70    [мадам, я, вчера, слышать, ваш, пение, о, что,...  negative      skip  \n",
       "71    [все, кто, предавать, прощать, все, кто, уходи...   neutral  positive  \n",
       "72                        [ужассс, извращёный, девочка]   neutral  negative  \n",
       "...                                                 ...       ...       ...  \n",
       "2893                           [любить, ложечка, сахар]   neutral  positive  \n",
       "2903  [приспичить, белок, покупать, под, конец, осен...   neutral      skip  \n",
       "2904  [вместо, лицо, аватары., .вместо, настроение, ...   neutral  negative  \n",
       "2905  [вятка, спасибо, за, великолепный, выходной, э...  positive    speech  \n",
       "2907  [поздравте., ., я, сегодня, племянница, радить...   neutral  positive  \n",
       "2908   [я, любить, свой, сестричка, вона, в, я, цьом**]  positive      skip  \n",
       "2909  [душа, хотеть, романтики., .поп, требовать, пр...  negative  positive  \n",
       "2911                        [вот, оно, искусство, ), )]   neutral  positive  \n",
       "2916  [с, святой, вы, святой, миколай, най, принес, ...    speech      skip  \n",
       "2917  [помнить, как, быть, тяжелый, но, очень, хотет...  positive  negative  \n",
       "2920                  [когда-нибудь, и, мы, пожениться]  positive      skip  \n",
       "2921  [я, прощать, одиночество, и, грусть, ты, сказа...  positive   neutral  \n",
       "2925             [иногда, увольнять, полезный, ), ), )]   neutral  positive  \n",
       "2931  [нуфвсё, гамовирать, учеба, я, теперь, маджист...   neutral  positive  \n",
       "2932  [сегодня, прогуляться, с, довольно, интересный...   neutral  negative  \n",
       "2934  [ты, чоооо-о, не, поздравлять, я, с, это, вели...    speech   neutral  \n",
       "2937  [многие, пиздить, в, инет, за, этот, фильм, ),...   neutral      skip  \n",
       "2940                                       [танюшечка❤]   neutral  positive  \n",
       "2942                                       [аркгаркгай]   neutral      skip  \n",
       "2944                    [завтра, в, я, перать, екзамен]   neutral      skip  \n",
       "2946  [стоить, смотреть, в, бесконечность, задумыват...  positive   neutral  \n",
       "2949                                        [тащуся, )]   neutral  positive  \n",
       "2951                                       [паздравляю]   neutral    speech  \n",
       "2954                                 [ващео, красаваа.]   neutral  positive  \n",
       "2955  [в, этот, песня, я, просто, влюбить, еще, с, 9...   neutral  positive  \n",
       "2957  [אימון, קרב, מגע, מתקדם, ביחידה, מובחרת, krav,...   neutral      skip  \n",
       "2959           [бля, я, и, забывать, какой, он, ахуать]  negative  positive  \n",
       "2961  [яркий, эмоция, незабываемый, встреча, и, поле...   neutral    speech  \n",
       "2964  [ты, один, половинка, я, второй, половинка, а,...  positive      skip  \n",
       "2966           [ахахахахы, прикольно, попереписуваться]  positive      skip  \n",
       "\n",
       "[1011 rows x 4 columns]"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "errors"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}