code/gijoong/03_tf_idf_option_test.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings(action='ignore')\n",
    "\n",
    "def kaggle_format(df):\n",
    "    df['label'][df['label'] == 'none'] = 0\n",
    "    df['label'][df['label'] == 'offensive'] = 1\n",
    "    df['label'][df['label'] == 'hate'] = 2\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "# train 데이터 / 7893/ \n",
    "# 'none'/ 'offensive'/ 'hate' : 3486/ 2498/ 1909\n",
    "train = pd.read_csv('total_20210121.csv')\n",
    "train = train[['comments', 'hate']]\n",
    "train.columns = ['comments', 'label']\n",
    "train = kaggle_format(train)\n",
    "train = train.astype({'label': 'str'})\n",
    "\n",
    "# dev 데이터 / 471/ \n",
    "# 'none'/ 'offensive'/ 'hate' : 160/ 189/ 122\n",
    "dev = pd.read_csv('./korean-hate-speech-master/labeled/dev.tsv', sep='\\t')\n",
    "dev = dev[['comments', 'hate']]\n",
    "dev.columns = ['comments', 'label']\n",
    "dev = kaggle_format(dev)\n",
    "dev = dev.astype({'label': 'str'})\n",
    "\n",
    "test = pd.read_csv('./korean-hate-speech-master/test.no_label.tsv', sep='\\t')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "X, y = train.comments, train.label\n",
    "X_test, y_test = dev.comments, dev.label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(7893, 146518)\n"
     ]
    }
   ],
   "source": [
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "\n",
    "vec = TfidfVectorizer(min_df=0.0, analyzer='char', ngram_range=(1,3), \n",
    "                      sublinear_tf=True, max_features=100000)\n",
    "X_tf = vec.fit_transform(X)\n",
    "print(X_tf.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "lgs = LogisticRegression(multi_class='multinomial', solver='lbfgs', C=1, \n",
    "                     class_weight='balanced', \n",
    "                     max_iter=6000, random_state=10)\n",
    "lgs.fit(X_tf, y)\n",
    "X_test_tf = vec.transform(X_test)\n",
    "pred =  lgs.predict(X_test_tf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6190162674997876"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import f1_score\n",
    "f1_score(y_test, pred, average='macro')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### TF-IDF Vectorizer\n",
    "- max_feature 값 변경 테스트 \n",
    "    - 최대 말뭉치 : 146518\n",
    "    - 값 : 10000 ~ 150000, 단위 10000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_f1_score(pre_data, mf):\n",
    "    vec = TfidfVectorizer(min_df=0.0, analyzer='char', ngram_range=(1,3), sublinear_tf=True,\n",
    "               max_features=mf)\n",
    "    X_tf = vec.fit_transform(pre_data)\n",
    "    print(X_tf.shape)\n",
    "    lgs = LogisticRegression(multi_class='multinomial', solver='lbfgs', C=1, \n",
    "                         class_weight='balanced', \n",
    "                         max_iter=6000, random_state=10)\n",
    "    lgs.fit(X_tf, y)\n",
    "    X_test_tf = vec.transform(X_test)\n",
    "    pred =  lgs.predict(X_test_tf)\n",
    "    score = f1_score(y_test, pred, average='macro')\n",
    "    return score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "score_compare = pd.DataFrame()\n",
    "mf_ls = list(range(10000, 150001, 10000))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(7893, 10000)\n",
      "(7893, 20000)\n",
      "(7893, 30000)\n",
      "(7893, 40000)\n",
      "(7893, 50000)\n",
      "(7893, 60000)\n",
      "(7893, 70000)\n",
      "(7893, 80000)\n",
      "(7893, 90000)\n",
      "(7893, 100000)\n",
      "(7893, 110000)\n",
      "(7893, 120000)\n",
      "(7893, 130000)\n",
      "(7893, 140000)\n",
      "(7893, 146518)\n"
     ]
    }
   ],
   "source": [
    "score_ls = []\n",
    "for f in mf_ls:\n",
    "    score = get_f1_score(X, f)\n",
    "    score_ls.append(score)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "score_compare['score'] = score_ls\n",
    "score_compare['max_feature'] = mf_ls "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "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>score</th>\n",
       "      <th>max_feature</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.591066</td>\n",
       "      <td>10000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.609017</td>\n",
       "      <td>20000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.610105</td>\n",
       "      <td>30000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.613206</td>\n",
       "      <td>40000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.613780</td>\n",
       "      <td>50000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.616511</td>\n",
       "      <td>60000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.620896</td>\n",
       "      <td>70000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.617150</td>\n",
       "      <td>80000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.615056</td>\n",
       "      <td>90000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.617150</td>\n",
       "      <td>100000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.608632</td>\n",
       "      <td>110000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.613861</td>\n",
       "      <td>120000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.616937</td>\n",
       "      <td>130000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.619016</td>\n",
       "      <td>140000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0.619016</td>\n",
       "      <td>150000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       score  max_feature\n",
       "0   0.591066        10000\n",
       "1   0.609017        20000\n",
       "2   0.610105        30000\n",
       "3   0.613206        40000\n",
       "4   0.613780        50000\n",
       "5   0.616511        60000\n",
       "6   0.620896        70000\n",
       "7   0.617150        80000\n",
       "8   0.615056        90000\n",
       "9   0.617150       100000\n",
       "10  0.608632       110000\n",
       "11  0.613861       120000\n",
       "12  0.616937       130000\n",
       "13  0.619016       140000\n",
       "14  0.619016       150000"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "score_compare"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "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>score</th>\n",
       "      <th>max_feature</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.620896</td>\n",
       "      <td>70000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.619016</td>\n",
       "      <td>140000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0.619016</td>\n",
       "      <td>150000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.617150</td>\n",
       "      <td>80000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.617150</td>\n",
       "      <td>100000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       score  max_feature\n",
       "6   0.620896        70000\n",
       "13  0.619016       140000\n",
       "14  0.619016       150000\n",
       "7   0.617150        80000\n",
       "9   0.617150       100000"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "score_compare.sort_values(by='score', ascending=False).head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_test(df, X_train, mf=100000, ngram=(1,3)):\n",
    "    vec = TfidfVectorizer(min_df=0.0, analyzer='char', ngram_range=ngram, sublinear_tf=True,\n",
    "               max_features=mf)\n",
    "    X_tf = vec.fit_transform(X_train)\n",
    "    print(X_tf.shape)\n",
    "    lgs = LogisticRegression(multi_class='multinomial', solver='lbfgs', C=1, \n",
    "                         class_weight='balanced', \n",
    "                         max_iter=6000, random_state=10).fit(X_tf, y)\n",
    "    X_test_tf = vec.transform(df['comments'])\n",
    "    df['label'] = lgs.predict(X_test_tf)\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(7893, 70000)\n"
     ]
    }
   ],
   "source": [
    "test = pd.read_csv('./korean-hate-speech-master/test.no_label.tsv', sep='\\t')\n",
    "test_df = get_test(test, X)\n",
    "test_df.to_csv('mf_70000.csv', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### ngram\n",
    "    - (2, 3) : (7893, 144860)\n",
    "    - (1, 4) : (7893, 337257)\n",
    "    - (2, 4) : (7893, 335599)\n",
    "    - (1, 5) : (7893, 572846)\n",
    "    - (2, 5) : (7893, 571188)\n",
    "    - (1, 6) : (7893, 824542)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_f1_score(pre_data, mf, ngram):\n",
    "    vec = TfidfVectorizer(min_df=0.0, analyzer='char', ngram_range=ngram, sublinear_tf=True,\n",
    "               max_features=mf)\n",
    "    X_tf = vec.fit_transform(pre_data)\n",
    "    print(X_tf.shape)\n",
    "    lgs = LogisticRegression(multi_class='multinomial', solver='lbfgs', C=1, \n",
    "                         class_weight='balanced', \n",
    "                         max_iter=6000, random_state=10)\n",
    "    lgs.fit(X_tf, y)\n",
    "    X_test_tf = vec.transform(X_test)\n",
    "    pred =  lgs.predict(X_test_tf)\n",
    "    score = f1_score(y_test, pred, average='macro')\n",
    "    return score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "ngr = [(2, 3), (1, 4), (2, 4), (1, 5), (2, 5), (1, 6)]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "mf_3 = list(range(70000, 150000, 10000))\n",
    "mf_4 = list(range(170000, 340000, 10000))\n",
    "mf_5 = list(range(290000, 580000, 10000))\n",
    "mf_6 = list(range(420000, 850000, 10000))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(2, 3)\n",
      "(7893, 70000)\n",
      "(7893, 80000)\n",
      "(7893, 90000)\n",
      "(7893, 100000)\n",
      "(7893, 110000)\n",
      "(7893, 120000)\n",
      "(7893, 130000)\n",
      "(7893, 140000)\n",
      "(1, 4)\n",
      "(7893, 170000)\n",
      "(7893, 180000)\n",
      "(7893, 190000)\n",
      "(7893, 200000)\n",
      "(7893, 210000)\n",
      "(7893, 220000)\n",
      "(7893, 230000)\n",
      "(7893, 240000)\n",
      "(7893, 250000)\n",
      "(7893, 260000)\n",
      "(7893, 270000)\n",
      "(7893, 280000)\n",
      "(7893, 290000)\n",
      "(7893, 300000)\n",
      "(7893, 310000)\n",
      "(7893, 320000)\n",
      "(7893, 330000)\n",
      "(2, 4)\n",
      "(7893, 170000)\n",
      "(7893, 180000)\n",
      "(7893, 190000)\n",
      "(7893, 200000)\n",
      "(7893, 210000)\n",
      "(7893, 220000)\n",
      "(7893, 230000)\n",
      "(7893, 240000)\n",
      "(7893, 250000)\n",
      "(7893, 260000)\n",
      "(7893, 270000)\n",
      "(7893, 280000)\n",
      "(7893, 290000)\n",
      "(7893, 300000)\n",
      "(7893, 310000)\n",
      "(7893, 320000)\n",
      "(7893, 330000)\n",
      "(1, 5)\n",
      "(7893, 290000)\n",
      "(7893, 300000)\n",
      "(7893, 310000)\n",
      "(7893, 320000)\n",
      "(7893, 330000)\n",
      "(7893, 340000)\n",
      "(7893, 350000)\n",
      "(7893, 360000)\n",
      "(7893, 370000)\n",
      "(7893, 380000)\n",
      "(7893, 390000)\n",
      "(7893, 400000)\n",
      "(7893, 410000)\n",
      "(7893, 420000)\n",
      "(7893, 430000)\n",
      "(7893, 440000)\n",
      "(7893, 450000)\n",
      "(7893, 460000)\n",
      "(7893, 470000)\n",
      "(7893, 480000)\n",
      "(7893, 490000)\n",
      "(7893, 500000)\n",
      "(7893, 510000)\n",
      "(7893, 520000)\n",
      "(7893, 530000)\n",
      "(7893, 540000)\n",
      "(7893, 550000)\n",
      "(7893, 560000)\n",
      "(7893, 570000)\n",
      "(2, 5)\n",
      "(7893, 290000)\n",
      "(7893, 300000)\n",
      "(7893, 310000)\n",
      "(7893, 320000)\n",
      "(7893, 330000)\n",
      "(7893, 340000)\n",
      "(7893, 350000)\n",
      "(7893, 360000)\n",
      "(7893, 370000)\n",
      "(7893, 380000)\n",
      "(7893, 390000)\n",
      "(7893, 400000)\n",
      "(7893, 410000)\n",
      "(7893, 420000)\n",
      "(7893, 430000)\n",
      "(7893, 440000)\n",
      "(7893, 450000)\n",
      "(7893, 460000)\n",
      "(7893, 470000)\n",
      "(7893, 480000)\n",
      "(7893, 490000)\n",
      "(7893, 500000)\n",
      "(7893, 510000)\n",
      "(7893, 520000)\n",
      "(7893, 530000)\n",
      "(7893, 540000)\n",
      "(7893, 550000)\n",
      "(7893, 560000)\n",
      "(7893, 570000)\n",
      "(1, 6)\n",
      "(7893, 420000)\n",
      "(7893, 430000)\n",
      "(7893, 440000)\n",
      "(7893, 450000)\n",
      "(7893, 460000)\n",
      "(7893, 470000)\n",
      "(7893, 480000)\n",
      "(7893, 490000)\n",
      "(7893, 500000)\n",
      "(7893, 510000)\n",
      "(7893, 520000)\n",
      "(7893, 530000)\n",
      "(7893, 540000)\n",
      "(7893, 550000)\n",
      "(7893, 560000)\n",
      "(7893, 570000)\n",
      "(7893, 580000)\n",
      "(7893, 590000)\n",
      "(7893, 600000)\n",
      "(7893, 610000)\n",
      "(7893, 620000)\n",
      "(7893, 630000)\n",
      "(7893, 640000)\n",
      "(7893, 650000)\n",
      "(7893, 660000)\n",
      "(7893, 670000)\n",
      "(7893, 680000)\n",
      "(7893, 690000)\n",
      "(7893, 700000)\n",
      "(7893, 710000)\n",
      "(7893, 720000)\n",
      "(7893, 730000)\n",
      "(7893, 740000)\n",
      "(7893, 750000)\n",
      "(7893, 760000)\n",
      "(7893, 770000)\n",
      "(7893, 780000)\n",
      "(7893, 790000)\n",
      "(7893, 800000)\n",
      "(7893, 810000)\n",
      "(7893, 820000)\n",
      "(7893, 824542)\n",
      "(7893, 824542)\n"
     ]
    }
   ],
   "source": [
    "score_ls = []\n",
    "ngram_ls = []\n",
    "mf_ls = []\n",
    "for ng in ngr:\n",
    "    print(ng)\n",
    "    if ng[1] == 3:\n",
    "        mf = mf_3\n",
    "    elif ng[1] == 4:\n",
    "        mf = mf_4\n",
    "    elif ng[1] == 5:\n",
    "        mf = mf_5\n",
    "    elif ng[1] == 6:\n",
    "        mf = mf_6\n",
    "    for f in mf:\n",
    "        score = get_f1_score(X, f, ng)\n",
    "        ngram_ls.append(ng)\n",
    "        mf_ls.append(f)\n",
    "        score_ls.append(score)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "ngram_df = pd.DataFrame()\n",
    "ngram_df['score'] = score_ls\n",
    "ngram_df['max_feature'] = mf_ls\n",
    "ngram_df['ngram_range'] = ngram_ls"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "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>score</th>\n",
       "      <th>max_feature</th>\n",
       "      <th>ngram_range</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.596407</td>\n",
       "      <td>70000</td>\n",
       "      <td>(2, 3)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.601048</td>\n",
       "      <td>80000</td>\n",
       "      <td>(2, 3)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.603002</td>\n",
       "      <td>90000</td>\n",
       "      <td>(2, 3)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.597915</td>\n",
       "      <td>100000</td>\n",
       "      <td>(2, 3)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.600517</td>\n",
       "      <td>110000</td>\n",
       "      <td>(2, 3)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>138</th>\n",
       "      <td>0.594967</td>\n",
       "      <td>800000</td>\n",
       "      <td>(1, 6)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>139</th>\n",
       "      <td>0.594967</td>\n",
       "      <td>810000</td>\n",
       "      <td>(1, 6)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>140</th>\n",
       "      <td>0.592731</td>\n",
       "      <td>820000</td>\n",
       "      <td>(1, 6)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>141</th>\n",
       "      <td>0.588531</td>\n",
       "      <td>830000</td>\n",
       "      <td>(1, 6)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>142</th>\n",
       "      <td>0.588531</td>\n",
       "      <td>840000</td>\n",
       "      <td>(1, 6)</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>143 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        score  max_feature ngram_range\n",
       "0    0.596407        70000      (2, 3)\n",
       "1    0.601048        80000      (2, 3)\n",
       "2    0.603002        90000      (2, 3)\n",
       "3    0.597915       100000      (2, 3)\n",
       "4    0.600517       110000      (2, 3)\n",
       "..        ...          ...         ...\n",
       "138  0.594967       800000      (1, 6)\n",
       "139  0.594967       810000      (1, 6)\n",
       "140  0.592731       820000      (1, 6)\n",
       "141  0.588531       830000      (1, 6)\n",
       "142  0.588531       840000      (1, 6)\n",
       "\n",
       "[143 rows x 3 columns]"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ngram_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {
    "scrolled": true
   },
   "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>score</th>\n",
       "      <th>max_feature</th>\n",
       "      <th>ngram_range</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>0.619642</td>\n",
       "      <td>240000</td>\n",
       "      <td>(1, 4)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0.619401</td>\n",
       "      <td>230000</td>\n",
       "      <td>(1, 4)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.619127</td>\n",
       "      <td>190000</td>\n",
       "      <td>(1, 4)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101</th>\n",
       "      <td>0.617765</td>\n",
       "      <td>430000</td>\n",
       "      <td>(1, 6)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.617755</td>\n",
       "      <td>180000</td>\n",
       "      <td>(1, 4)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>0.617579</td>\n",
       "      <td>250000</td>\n",
       "      <td>(1, 4)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.617455</td>\n",
       "      <td>170000</td>\n",
       "      <td>(1, 4)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.617263</td>\n",
       "      <td>220000</td>\n",
       "      <td>(1, 4)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.617057</td>\n",
       "      <td>200000</td>\n",
       "      <td>(1, 4)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>0.617010</td>\n",
       "      <td>280000</td>\n",
       "      <td>(1, 4)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>0.616672</td>\n",
       "      <td>320000</td>\n",
       "      <td>(1, 4)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0.616672</td>\n",
       "      <td>300000</td>\n",
       "      <td>(1, 4)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>0.616672</td>\n",
       "      <td>290000</td>\n",
       "      <td>(1, 4)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>0.616356</td>\n",
       "      <td>420000</td>\n",
       "      <td>(1, 5)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>0.616045</td>\n",
       "      <td>330000</td>\n",
       "      <td>(1, 5)</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        score  max_feature ngram_range\n",
       "15   0.619642       240000      (1, 4)\n",
       "14   0.619401       230000      (1, 4)\n",
       "10   0.619127       190000      (1, 4)\n",
       "101  0.617765       430000      (1, 6)\n",
       "9    0.617755       180000      (1, 4)\n",
       "16   0.617579       250000      (1, 4)\n",
       "8    0.617455       170000      (1, 4)\n",
       "13   0.617263       220000      (1, 4)\n",
       "11   0.617057       200000      (1, 4)\n",
       "19   0.617010       280000      (1, 4)\n",
       "23   0.616672       320000      (1, 4)\n",
       "21   0.616672       300000      (1, 4)\n",
       "20   0.616672       290000      (1, 4)\n",
       "55   0.616356       420000      (1, 5)\n",
       "46   0.616045       330000      (1, 5)"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ngram_df.sort_values(by='score', ascending=False).head(15)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(7893, 190000)\n"
     ]
    }
   ],
   "source": [
    "test = pd.read_csv('./korean-hate-speech-master/test.no_label.tsv', sep='\\t')\n",
    "test_df = get_test(test, X, 190000, ngram=(1,4))\n",
    "test_df.to_csv('ngram_14_mf_190000.csv', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}