text_processing.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "pd.set_option(\"display.max_columns\", None)\n",
    "pd.set_option(\"display.max_rows\", None)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Para Classifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('TXT/Li_paragraph.txt', 'r') as f:\n",
    "        paragraphs = f.readlines()\n",
    "\n",
    "with open('TXT/Li_label.txt', 'r') as f:\n",
    "        labels = f.readlines()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def transform_user_features(paragraph):\n",
    "    \n",
    "    numbers = sum(c.isdigit() for c in paragraph)\n",
    "    letters = sum(c.isalpha() for c in paragraph)\n",
    "\n",
    "    features = [\n",
    "      #Character length thresholds\n",
    "      bool(len(paragraph) > 10),\n",
    "      bool(len(paragraph) > 100),\n",
    "      bool(len(paragraph) > 1000),\n",
    "      bool(len(paragraph) > 10000),\n",
    "\n",
    "      #Number and word counts\n",
    "      bool(numbers >= 1),\n",
    "      bool(numbers > 10),\n",
    "      bool(numbers > 100),\n",
    "      bool(letters >= 1),\n",
    "      bool(letters > 10),\n",
    "      bool(letters > 100),\n",
    "      bool(letters > 1000),\n",
    "      bool(letters > 10000),\n",
    "\n",
    "      #Special characters\n",
    "      bool( sum(c in ['(', ')'] for c in paragraph) >= 1),\n",
    "      bool( sum(c in ['(', ')'] for c in paragraph) >= 10),\n",
    "      bool( sum(c in ['(', ')'] for c in paragraph) >= 100),\n",
    "      bool( sum(c in ['[', ']'] for c in paragraph) >= 1),\n",
    "      bool( sum(c in ['[', ']'] for c in paragraph) >= 10),\n",
    "      bool( sum(c in ['[', ']'] for c in paragraph) >= 100),\n",
    "\n",
    "      #Heading heuristics\n",
    "      bool('experiment' in paragraph[:50].lower()),\n",
    "      bool('synthesi' in paragraph[:50].lower()),\n",
    "      bool('prepar' in paragraph[:50].lower()),\n",
    "      bool('abstract' in paragraph[:50].lower()),\n",
    "      bool('characteriz' in paragraph[:50].lower()),\n",
    "\n",
    "      #Domain keyword checks\n",
    "      bool('heat' in paragraph.lower()),\n",
    "      bool('dissolve' in paragraph.lower()),\n",
    "      bool('mix' in paragraph.lower()),\n",
    "      bool('obtain' in paragraph.lower()),\n",
    "\n",
    "      bool('mol ' in paragraph.lower()),\n",
    "      bool('%' in paragraph.lower()),\n",
    "      bool('ml ' in paragraph.lower()),\n",
    "      bool(' ph ' in paragraph.lower()),\n",
    "\n",
    "      bool('ratio' in paragraph.lower()),\n",
    "      bool('stoichiometric' in paragraph.lower()),\n",
    "\n",
    "      bool('sample' in paragraph.lower()),\n",
    "      bool('solution' in paragraph.lower()),\n",
    "      bool('product' in paragraph.lower()),\n",
    "      bool('chemical' in paragraph.lower()),\n",
    "\n",
    "      bool('study' in paragraph.lower()),\n",
    "      bool('method' in paragraph.lower()),\n",
    "      bool('technique' in paragraph.lower()),\n",
    "      bool('route' in paragraph.lower()),\n",
    "\n",
    "      #Heuristic phrases\n",
    "      bool('was prepared by' in paragraph.lower()),\n",
    "      bool('dissolved in' in paragraph.lower()),\n",
    "      bool('final product' in paragraph.lower()),\n",
    "      bool('the precursors' in paragraph.lower()),\n",
    "      bool('purchased from' in paragraph.lower()),\n",
    "\n",
    "      #Characterization\n",
    "      bool('xrd' in paragraph.lower()),\n",
    "      bool('ftir' in paragraph.lower()),\n",
    "      bool('voltammetry' in paragraph.lower()),\n",
    "      bool('sem ' in paragraph.lower()),\n",
    "      bool('microscop' in paragraph.lower()),\n",
    "      bool('spectroscop' in paragraph.lower()),\n",
    "    ]\n",
    "\n",
    "    return np.array(features)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "paragraphs = list(map(lambda p: p.split('\\n')[0], paragraphs))\n",
    "labels = list(map(lambda l: int(l.split('\\n')[0]), labels))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "\n",
    "vectorizer = CountVectorizer(max_features=100)\n",
    "vectorizer.fit(paragraphs)\n",
    "cv_vectors  = vectorizer.transform(paragraphs).toarray()\n",
    "ud_vectors = np.array([transform_user_features(p) for p in paragraphs])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1008, 152)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "features = np.hstack([ud_vectors, cv_vectors])\n",
    "features.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "labels = np.array(labels)\n",
    "\n",
    "features = features[np.where(labels<2,)]\n",
    "labels = labels[np.where(labels<2,)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.svm import SVC\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(features, labels, test_size=0.2, random_state=96)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/ynyuan/anaconda3/envs/nlp/lib/python3.6/site-packages/sklearn/svm/base.py:193: FutureWarning: The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "  \"avoid this warning.\", FutureWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.9352941176470588"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "reg = SVC(C=0.1, probability=True, random_state=120)\n",
    "reg.fit(X_train, y_train)\n",
    "reg.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.metrics import roc_curve, roc_auc_score\n",
    "\n",
    "plt.style.use('seaborn')\n",
    "\n",
    "plt.plot(roc_curve(y_train, reg.predict_proba(X_train)[:,1])[0], roc_curve(y_train, reg.predict_proba(X_train)[:,1])[1], \n",
    "         color='blue', label='Train ROC Curve (area = %0.2f)' % roc_auc_score(y_train, reg.predict_proba(X_train)[:,1]))\n",
    "\n",
    "plt.plot(roc_curve(y_test, reg.predict_proba(X_test)[:,1])[0], roc_curve(y_test, reg.predict_proba(X_test)[:,1])[1], \n",
    "         color='red', label='Test ROC Curve (area = %0.2f)' % roc_auc_score(y_test, reg.predict_proba(X_test)[:,1]))\n",
    "\n",
    "\n",
    "plt.plot([0, 1], [0, 1], color='black', linestyle='--')\n",
    "plt.xlim([0.0, 1.0])\n",
    "plt.ylim([0.0, 1.05])\n",
    "plt.title(\"ROC Curve of Paragraph Classifier\")\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.legend()\n",
    "#plt.show()\n",
    "plt.savefig('para_classifier_roc')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Word Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "label_dict = {0:'null', 1:'reactant', 2:'product', 3:'genetic material', 4:'condition',\n",
    "              5:'apparatus', 6:'operation', 7:'synthesis category', 8:'condition unit', 9:'amount unit',\n",
    "              10:'number', 11:'condition description', 12:'amount description'}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "11"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "with open('TXT/text3.txt', 'r') as f:\n",
    "        text = f.readlines()\n",
    "text = list(map(lambda p: p.split('\\n')[0], text))\n",
    "len(text)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Zinc-doped MgO (ZnxMg1−xO) powders were synthesized in a flame spray reactor. Detailed description of the setup can be found elsewhere. For the preparation of the precursor solutions, 56.8 g of Mg(II) acetylacetonate was dissolved in a mixture of 2−ethylhexanoic acid (167.2 mL), ethanol absolute (100.0 mL) and certain amount of xylene, to make a final volume of 500.0 mL, corresponding to an Mg(II) concentration of 0.5 mol/L. The Zn(II) solution (0.5 mol/L) was made by diluting 81.7 g of Zn(II) naphthenate (in naphthenic acid as received) with xylene to a final volume of 250.0 mL. The Mg(II) and Zn(II) solutions were mixed at chosen volume ratio to achieve the desired Zn(II) doping levels while keeping the total metal concentration of 0.5 mol/L in the precursor solutions. During flame spray pyrolysis experiments, the mixed precursor solution was fed by a syringe pump (KD Scientific) at a fixed flow rate of 5.0 mL/min. The liquid was dispersed into fine droplets by O2 gas (5.0 L/min, 1.5 bar) at the spray nozzle exit. The spray was ignited by a CH4/O2 (1.5 and 3.2 L/min, respectively) supporting flame to form a self−sustained flame. The flow rates of all gases were controlled by calibrated mass flow controllers (Bronkhorst High-Tech). The generated ZnxMg1−xO particles were directed by a vacuum pump and collected on a water-cooled glass fiber filter (Whatman GF−6) placed 450 mm above the nozzle. After being collected from the filter surface, the powders were sieved to remove the glass fibers.'"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "text[3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "from nltk.corpus import stopwords\n",
    "\n",
    "stop = stopwords.words('english')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_words = []\n",
    "for para in text: \n",
    "    tokens = para.split(' ')\n",
    "    tokens = [word.strip(\"\\(\\).\\),\\[\\].∼;\") for word in tokens]\n",
    "    #tokens = [word for word in tokens if word not in stop]\n",
    "    all_words.extend(tokens)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(2105, 1)"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_words_df = pd.DataFrame(all_words, columns=['word'])\n",
    "all_words_df.shape\n",
    "#all_words_df.to_csv(\"words1_new2.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "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>word</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Synthesis</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Co−P</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Nanocomposites</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Co−P</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>nanocomposites</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             word  label\n",
       "0       Synthesis      7\n",
       "1            Co−P      2\n",
       "2  Nanocomposites      3\n",
       "3            Co−P      2\n",
       "4  nanocomposites      3"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "words1_df = pd.read_csv(\"CSV/words1_new_finish.csv\", index_col=0)\n",
    "words2_df = pd.read_csv(\"CSV/words2_new_labeled.csv\", index_col=0)\n",
    "words3_df = pd.read_csv(\"CSV/words3.csv\", index_col=0)\n",
    "words_df = pd.concat([words1_df, words2_df, words3_df], axis=0)\n",
    "words_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Synthesis',\n",
       " 'Co−P',\n",
       " 'Nanocomposites',\n",
       " 'Co−P',\n",
       " 'nanocomposites',\n",
       " 'fabricated',\n",
       " 'following',\n",
       " 'previous',\n",
       " 'literature',\n",
       " 'slight',\n",
       " 'modiﬁcations',\n",
       " 'In',\n",
       " 'detail',\n",
       " '1.2',\n",
       " 'mmol',\n",
       " 'Co(ac)2',\n",
       " 'mixed',\n",
       " '2.4',\n",
       " 'mmol',\n",
       " 'TPP',\n",
       " '21',\n",
       " 'mL',\n",
       " 'OM',\n",
       " '100',\n",
       " 'mL',\n",
       " 'three-necked-ﬂask',\n",
       " 'The',\n",
       " 'mixture',\n",
       " 'heated',\n",
       " '110',\n",
       " '°C',\n",
       " '30',\n",
       " 'min',\n",
       " 'argon-ﬂow',\n",
       " 'exclude',\n",
       " 'water',\n",
       " 'well',\n",
       " 'oxygen',\n",
       " 'heated',\n",
       " '290',\n",
       " '°C',\n",
       " '3',\n",
       " 'h',\n",
       " 'Co−P',\n",
       " 'nanocomposites',\n",
       " 'obtained',\n",
       " 'centrifugation',\n",
       " 'black',\n",
       " 'solution',\n",
       " 'cooled',\n",
       " 'room',\n",
       " 'temperature',\n",
       " 'Then',\n",
       " 'Co−P',\n",
       " 'nanocomposites',\n",
       " 'washed',\n",
       " 'mixed',\n",
       " 'solution',\n",
       " 'cyclohexane',\n",
       " 'acetone',\n",
       " 'ethanol',\n",
       " 'several',\n",
       " 'times',\n",
       " 'dispersed',\n",
       " '30',\n",
       " 'mL',\n",
       " 'cyclohexane',\n",
       " 'Catalyst',\n",
       " 'Synthesis',\n",
       " 'ZnFe2O4',\n",
       " 'prepared',\n",
       " 'according',\n",
       " 'previously',\n",
       " 'reported',\n",
       " 'method',\n",
       " 'CuOx-decorated',\n",
       " 'ZnFe2O4',\n",
       " 'prepared',\n",
       " 'replacing',\n",
       " 'Zn',\n",
       " 'ions',\n",
       " 'Cu',\n",
       " 'ions',\n",
       " 'All',\n",
       " 'analytic',\n",
       " 'reagents',\n",
       " 'dissolved',\n",
       " 'deionized',\n",
       " 'water',\n",
       " 'Typically',\n",
       " 'NaBH4',\n",
       " 'solution',\n",
       " 'dropwise',\n",
       " 'added',\n",
       " 'mixed',\n",
       " 'solution',\n",
       " 'Fe(NO3)3',\n",
       " 'Cu(NO3)2',\n",
       " 'Zn(NO3)2',\n",
       " 'Fe3+]/[NaBH4',\n",
       " '=',\n",
       " '1:2',\n",
       " 'After',\n",
       " 'stirring',\n",
       " '1',\n",
       " 'h',\n",
       " 'mixture',\n",
       " 'heated',\n",
       " '120',\n",
       " '°C',\n",
       " '12',\n",
       " 'h',\n",
       " 'autoclave',\n",
       " 'tube',\n",
       " 'The',\n",
       " 'solid',\n",
       " 'product',\n",
       " 'washed',\n",
       " 'deionized',\n",
       " 'water',\n",
       " 'ﬁnally',\n",
       " 'dried',\n",
       " '60',\n",
       " '°C',\n",
       " 'overnight',\n",
       " 'The',\n",
       " 'samples',\n",
       " 'prepared',\n",
       " 'different',\n",
       " 'Cu/Zn',\n",
       " 'mole',\n",
       " 'ratios',\n",
       " '1:9',\n",
       " '3:7',\n",
       " '5:5',\n",
       " '7:3',\n",
       " 'designated',\n",
       " 'ZnFe2O4−Cu1',\n",
       " 'ZnFe2O4−Cu3',\n",
       " 'ZnFe2O4−Cu5',\n",
       " 'ZnFe2O4−Cu7',\n",
       " 'As',\n",
       " 'reference',\n",
       " 'Cu−Fe',\n",
       " 'composite',\n",
       " 'oxide',\n",
       " 'designated',\n",
       " 'Cu−Fe−O',\n",
       " 'prepared',\n",
       " 'absence',\n",
       " 'Zn(NO3)2',\n",
       " 'Cu2+]/[Fe3+',\n",
       " '=',\n",
       " '1:2',\n",
       " 'In',\n",
       " 'typical',\n",
       " 'synthetic',\n",
       " 'procedure',\n",
       " 'composite',\n",
       " 'microspheres',\n",
       " 'pyrite',\n",
       " 'polyvinylpyrrolidone',\n",
       " 'PVP',\n",
       " 'ﬁrst',\n",
       " 'synthesized',\n",
       " 'brieﬂy',\n",
       " '0.556',\n",
       " 'g',\n",
       " 'FeSO4·7H2O',\n",
       " '0.6',\n",
       " 'g',\n",
       " 'PVP-K30',\n",
       " 'dissolved',\n",
       " '20',\n",
       " 'mL',\n",
       " 'ethylene',\n",
       " 'glycol',\n",
       " '100',\n",
       " 'mL',\n",
       " 'round-bottomed-ﬂask',\n",
       " 'Then',\n",
       " '0.128',\n",
       " 'g',\n",
       " 'S',\n",
       " 'powder',\n",
       " 'dispersed',\n",
       " 'solution',\n",
       " 'ultrasonication',\n",
       " 'minutes',\n",
       " 'The',\n",
       " 'round-bottomed-ﬂask',\n",
       " 'equipped',\n",
       " 'microwave',\n",
       " 'reactor',\n",
       " 'reaction',\n",
       " '30',\n",
       " 'min',\n",
       " 'nitrogen',\n",
       " 'ﬂow',\n",
       " '80%',\n",
       " 'full',\n",
       " 'power',\n",
       " 'black',\n",
       " 'precipitate',\n",
       " 'obtained',\n",
       " 'The',\n",
       " 'dried',\n",
       " 'black',\n",
       " 'precipitate',\n",
       " 'calcined',\n",
       " 'air',\n",
       " '900',\n",
       " '°C',\n",
       " '10',\n",
       " 'min',\n",
       " 'heating',\n",
       " 'rate',\n",
       " '5',\n",
       " '°C min−1',\n",
       " 'cooled',\n",
       " 'room',\n",
       " 'temperature',\n",
       " 'naturally',\n",
       " 'Finally',\n",
       " 'red',\n",
       " 'powder',\n",
       " 'obtained',\n",
       " 'collected',\n",
       " 'characterizations',\n",
       " 'Synthesis',\n",
       " 'GO',\n",
       " 'Na3V2(PO4)3',\n",
       " 'GO',\n",
       " 'prepared',\n",
       " 'improved',\n",
       " 'Hummers',\n",
       " 'course',\n",
       " 'The',\n",
       " 'resultant',\n",
       " 'GO',\n",
       " 'dispersion',\n",
       " 'underwent',\n",
       " 'dialysis',\n",
       " 'room',\n",
       " 'temperature',\n",
       " '14',\n",
       " 'days',\n",
       " 'rotary',\n",
       " 'evaporation',\n",
       " '60',\n",
       " '°C',\n",
       " 'remove',\n",
       " 'suitable',\n",
       " 'water',\n",
       " 'The',\n",
       " 'concentration',\n",
       " 'ﬁnal',\n",
       " 'GO',\n",
       " 'dispersion',\n",
       " 'adopted',\n",
       " '5.24',\n",
       " 'mg ',\n",
       " 'mL−1',\n",
       " 'In',\n",
       " 'typical',\n",
       " 'synthesis',\n",
       " 'Na3V2(PO4)3',\n",
       " 'NVP',\n",
       " 'NH4VO3',\n",
       " '2',\n",
       " 'mmol',\n",
       " 'oxalic',\n",
       " 'acid',\n",
       " '8.4',\n",
       " 'mmol',\n",
       " 'added',\n",
       " '80',\n",
       " 'mL',\n",
       " 'water',\n",
       " '80',\n",
       " '°C',\n",
       " 'After',\n",
       " 'stirring',\n",
       " '1',\n",
       " 'h',\n",
       " 'GO',\n",
       " 'dispersion',\n",
       " 'added',\n",
       " 'mixture',\n",
       " 'sonication',\n",
       " '0.5',\n",
       " 'h',\n",
       " 'Then',\n",
       " 'Na2CO3',\n",
       " '1.5',\n",
       " 'mmol',\n",
       " 'NH4H2PO4',\n",
       " '3',\n",
       " 'mmol',\n",
       " 'added',\n",
       " 'agitation',\n",
       " '36',\n",
       " 'h',\n",
       " 'The',\n",
       " 'resultant',\n",
       " 'product',\n",
       " 'precalcined',\n",
       " '350',\n",
       " '°C',\n",
       " '3',\n",
       " 'h',\n",
       " 'N2',\n",
       " 'atmosphere',\n",
       " 'followed',\n",
       " 'powder',\n",
       " 'produced',\n",
       " 'calcined',\n",
       " '700',\n",
       " '°C',\n",
       " '10',\n",
       " 'h',\n",
       " 'H2−Ar',\n",
       " 'mixture',\n",
       " '5%',\n",
       " 'H2',\n",
       " 'v/v',\n",
       " 'CuInSe2',\n",
       " 'Nanocrystal',\n",
       " 'Synthesis',\n",
       " 'CIS',\n",
       " 'nanocrystals',\n",
       " 'synthesized',\n",
       " 'using',\n",
       " 'modiﬁcation',\n",
       " 'published',\n",
       " 'procedures',\n",
       " 'In',\n",
       " 'typical',\n",
       " 'reaction',\n",
       " '2',\n",
       " 'mmol',\n",
       " 'CuCl',\n",
       " '2',\n",
       " 'mmol',\n",
       " 'InCl3',\n",
       " '4',\n",
       " 'mmol',\n",
       " 'Se',\n",
       " '20',\n",
       " 'mL',\n",
       " 'degassed',\n",
       " 'oleylamine',\n",
       " 'added',\n",
       " '100',\n",
       " 'mL',\n",
       " 'three-neck-ﬂask',\n",
       " 'inside',\n",
       " 'N2',\n",
       " 'ﬁlled',\n",
       " 'glovebox',\n",
       " 'The',\n",
       " 'ﬂask',\n",
       " 'attached',\n",
       " 'standard',\n",
       " 'Schlenk',\n",
       " 'line',\n",
       " 'degassed',\n",
       " '110',\n",
       " '°C',\n",
       " 'vacuum',\n",
       " '45',\n",
       " 'min',\n",
       " 'The',\n",
       " 'ﬂask',\n",
       " 'ﬁlled',\n",
       " 'nitrogen',\n",
       " 'heated',\n",
       " '200',\n",
       " '°C',\n",
       " 'After',\n",
       " '30',\n",
       " 'min',\n",
       " 'temperature',\n",
       " 'raised',\n",
       " '260',\n",
       " '°C',\n",
       " 'After',\n",
       " '10',\n",
       " 'min',\n",
       " 'heating',\n",
       " 'mantle',\n",
       " 'removed',\n",
       " 'reaction',\n",
       " 'allowed',\n",
       " 'cool',\n",
       " 'room',\n",
       " 'temperature',\n",
       " 'The',\n",
       " 'nanocrystals',\n",
       " 'washed',\n",
       " 'via',\n",
       " 'centrifugation',\n",
       " 'using',\n",
       " 'toluene',\n",
       " 'ethanol',\n",
       " 'solvent',\n",
       " 'antisolvent',\n",
       " 'respectively',\n",
       " 'Poorly',\n",
       " 'capped',\n",
       " 'nanocrystals',\n",
       " 'separated',\n",
       " 'toluene',\n",
       " 'solution',\n",
       " 'via',\n",
       " 'centrifugation',\n",
       " 'The',\n",
       " 'ﬁnal',\n",
       " 'nanocrystal',\n",
       " 'dispersion',\n",
       " 'transferred',\n",
       " 'glovebox',\n",
       " 'ligand',\n",
       " 'exchange',\n",
       " 'Synthesis',\n",
       " 'PEI-PTMC',\n",
       " 'PTMC20k-COOH',\n",
       " '4.0',\n",
       " 'g',\n",
       " '0.2',\n",
       " 'mmol',\n",
       " 'DCC',\n",
       " '4.1',\n",
       " 'g',\n",
       " '20',\n",
       " 'mmol',\n",
       " 'dissolved',\n",
       " '10',\n",
       " 'mL',\n",
       " 'dried',\n",
       " 'DMF',\n",
       " 'stirred',\n",
       " 'room',\n",
       " 'temperature',\n",
       " '2',\n",
       " 'h',\n",
       " 'Then',\n",
       " 'PEI25K',\n",
       " '5.0',\n",
       " 'g',\n",
       " '0.2',\n",
       " 'mmol',\n",
       " 'DMAP',\n",
       " '0.27',\n",
       " 'g',\n",
       " '2',\n",
       " 'mmol',\n",
       " 'added',\n",
       " 'solution',\n",
       " 'The',\n",
       " 'solution',\n",
       " 'kept',\n",
       " 'stirring',\n",
       " 'room',\n",
       " 'temperature',\n",
       " '24',\n",
       " 'h',\n",
       " 'The',\n",
       " 'resulting',\n",
       " 'mixture',\n",
       " 'ﬁltered',\n",
       " 'ﬁltrate',\n",
       " 'dialyzed',\n",
       " 'MWCO',\n",
       " '=',\n",
       " '30000',\n",
       " 'Da',\n",
       " 'distilled',\n",
       " 'water',\n",
       " '48',\n",
       " 'h',\n",
       " 'lyophilized',\n",
       " 'obtain',\n",
       " 'PEI-PTMC',\n",
       " 'Synthesis',\n",
       " 'PEG-DA',\n",
       " 'mPEG2k-NH2',\n",
       " '400',\n",
       " 'mg',\n",
       " '0.2',\n",
       " 'mmol',\n",
       " 'DA',\n",
       " '276',\n",
       " 'mg',\n",
       " '2',\n",
       " 'mmol',\n",
       " '20',\n",
       " 'mL',\n",
       " 'PBS',\n",
       " 'pH',\n",
       " '8.0',\n",
       " 'stirred',\n",
       " 'room',\n",
       " 'temperature',\n",
       " '24',\n",
       " 'h',\n",
       " 'Then',\n",
       " 'solution',\n",
       " 'dialyzed',\n",
       " 'MWCO',\n",
       " '=',\n",
       " '1000',\n",
       " 'Da',\n",
       " 'distilled',\n",
       " 'water',\n",
       " 'neutral',\n",
       " 'pH',\n",
       " '24',\n",
       " 'h',\n",
       " 'lyophilized',\n",
       " 'obtain',\n",
       " 'PEG-DA',\n",
       " '°C',\n",
       " 'Synthesis',\n",
       " 'LiMgxCuyCo1−x−yO2',\n",
       " 'carried',\n",
       " 'mixing',\n",
       " 'stoichiometric',\n",
       " 'amounts',\n",
       " 'Li(COOCH3)2·2H2O',\n",
       " 'Co(COOCH3)2·4H2O',\n",
       " 'Cu(NO3)2·3H2O',\n",
       " 'Mg(COOCH3)2·4H2O',\n",
       " 'dissolving',\n",
       " 'minimum',\n",
       " 'quantity',\n",
       " 'triple',\n",
       " 'distilled',\n",
       " 'water',\n",
       " 'The',\n",
       " 'resulting',\n",
       " 'solution',\n",
       " 'concentrated',\n",
       " 'stirring',\n",
       " 'continuously',\n",
       " '100',\n",
       " 'Thus',\n",
       " 'concentrated',\n",
       " 'solution',\n",
       " 'transferred',\n",
       " 'silica',\n",
       " 'crucible',\n",
       " 'placed',\n",
       " 'centre',\n",
       " 'rotating',\n",
       " 'plate',\n",
       " 'microwave',\n",
       " 'oven',\n",
       " 'Kenstar',\n",
       " 'India',\n",
       " '2450',\n",
       " 'MHz',\n",
       " '800',\n",
       " 'W',\n",
       " 'The',\n",
       " 'solution',\n",
       " 'irradiated',\n",
       " '100%',\n",
       " 'power',\n",
       " '15−25',\n",
       " 'minutes',\n",
       " 'During',\n",
       " 'reaction',\n",
       " 'acetates',\n",
       " 'nitrates',\n",
       " 'quickly',\n",
       " 'burnt',\n",
       " 'red',\n",
       " 'glow',\n",
       " 'appears',\n",
       " 'inside',\n",
       " 'silica',\n",
       " 'crucible',\n",
       " 'throughout',\n",
       " 'reaction',\n",
       " 'After',\n",
       " 'irradiation',\n",
       " 'product',\n",
       " 'ground',\n",
       " 'hours',\n",
       " 'obtain',\n",
       " 'phase',\n",
       " 'pure',\n",
       " 'LiMgxCuyCo1−x−yO2',\n",
       " 'A',\n",
       " 'Teflon-lined',\n",
       " 'stainless',\n",
       " 'steel',\n",
       " 'autoclave',\n",
       " 'filled',\n",
       " '20',\n",
       " 'mL',\n",
       " 'aqueous',\n",
       " 'solution',\n",
       " 'containing',\n",
       " '0.1',\n",
       " 'M',\n",
       " 'FeCl3·6H2O',\n",
       " '1',\n",
       " 'M',\n",
       " 'NaNO3',\n",
       " 'A',\n",
       " 'piece',\n",
       " 'FTO',\n",
       " 'glass',\n",
       " 'cleaned',\n",
       " 'ultrasonically',\n",
       " 'pure',\n",
       " 'water',\n",
       " 'ethanol',\n",
       " 'acetone',\n",
       " 'rinsed',\n",
       " 'water',\n",
       " 'chemical',\n",
       " 'bath',\n",
       " 'deposition',\n",
       " 'Then',\n",
       " 'FTO',\n",
       " 'glass',\n",
       " 'put',\n",
       " 'autoclave',\n",
       " 'heated',\n",
       " '95',\n",
       " '°C',\n",
       " 'A',\n",
       " 'uniform',\n",
       " 'yellow',\n",
       " 'layer',\n",
       " 'iron',\n",
       " 'oxyhydrodides',\n",
       " 'FeOOH',\n",
       " 'film',\n",
       " 'formed',\n",
       " 'FTO',\n",
       " 'substrate',\n",
       " 'cooled',\n",
       " 'slowly',\n",
       " 'room',\n",
       " 'temperature',\n",
       " 'PEC',\n",
       " 'measurements',\n",
       " 'In',\n",
       " 'different',\n",
       " 'approach',\n",
       " 'hematite-nanostructured',\n",
       " 'photoanodes',\n",
       " 'prepared',\n",
       " 'conditions',\n",
       " 'except',\n",
       " 'lack',\n",
       " 'using',\n",
       " 'inorganic',\n",
       " 'salt',\n",
       " 'NaNO3',\n",
       " 'electrolyte',\n",
       " 'process',\n",
       " 'chemical',\n",
       " 'bath',\n",
       " 'deposition',\n",
       " 'Vertically',\n",
       " 'aligned',\n",
       " 'Si',\n",
       " 'nanostructure',\n",
       " 'arrays',\n",
       " 'SiNs',\n",
       " 'boron-doped',\n",
       " 'p-type',\n",
       " 'single-side-polished',\n",
       " 'Si(100',\n",
       " 'substrates',\n",
       " 'resistivity',\n",
       " '=',\n",
       " '2−10',\n",
       " 'Ω',\n",
       " 'cm',\n",
       " 'fabricated',\n",
       " 'ICP-RIE',\n",
       " 'etching',\n",
       " 'technique',\n",
       " 'using',\n",
       " 'silver',\n",
       " 'nano-particles',\n",
       " 'mask',\n",
       " 'Si',\n",
       " 'substrates',\n",
       " 'first',\n",
       " 'cleaned',\n",
       " 'standard',\n",
       " 'RCA',\n",
       " 'cleaning',\n",
       " 'process',\n",
       " 'Thereafter',\n",
       " 'uniform',\n",
       " 'layer',\n",
       " 'Ag',\n",
       " 'nanoparticles',\n",
       " 'Ag',\n",
       " 'NPs',\n",
       " 'cleaned',\n",
       " 'Si',\n",
       " 'prepared',\n",
       " 'depositing',\n",
       " 'Ag',\n",
       " 'film',\n",
       " '10',\n",
       " 'nm',\n",
       " 'thick',\n",
       " 'thermal',\n",
       " 'evaporation',\n",
       " 'followed',\n",
       " 'annealing',\n",
       " 'inert',\n",
       " 'atmosphere',\n",
       " '400',\n",
       " '°C',\n",
       " '1',\n",
       " 'h',\n",
       " 'Si',\n",
       " 'pieces',\n",
       " '2',\n",
       " '×',\n",
       " '2',\n",
       " 'cm2',\n",
       " 'size',\n",
       " 'coated',\n",
       " 'Ag',\n",
       " 'NPs',\n",
       " 'average',\n",
       " 'diameter',\n",
       " '50',\n",
       " 'nm',\n",
       " 'transferred',\n",
       " 'SENTECH',\n",
       " 'SI',\n",
       " '500',\n",
       " 'ICP-RIE',\n",
       " 'system',\n",
       " 'employing',\n",
       " 'CF4',\n",
       " 'gas',\n",
       " 'etch',\n",
       " 'unmasked',\n",
       " 'area',\n",
       " 'In',\n",
       " 'CF4-based',\n",
       " 'plasma',\n",
       " 'etching',\n",
       " 'system',\n",
       " 'application',\n",
       " 'electric',\n",
       " 'field',\n",
       " 'free',\n",
       " 'electrons',\n",
       " 'collide',\n",
       " 'neutral',\n",
       " 'molecules',\n",
       " 'CF4',\n",
       " 'resulting',\n",
       " 'dissociation',\n",
       " 'CF4',\n",
       " 'electrons',\n",
       " 'ions',\n",
       " 'fluorine',\n",
       " 'radicals',\n",
       " 'dissociative',\n",
       " 'ionization',\n",
       " 'resulting',\n",
       " 'formation',\n",
       " 'volatile',\n",
       " 'SiF4',\n",
       " 'Though',\n",
       " 'dry',\n",
       " 'etching',\n",
       " 'combination',\n",
       " 'physical',\n",
       " 'chemical',\n",
       " 'mechanisms',\n",
       " 'outer',\n",
       " 'Si',\n",
       " 'atoms',\n",
       " 'dominantly',\n",
       " 'removed',\n",
       " 'chemical',\n",
       " 'reaction',\n",
       " 'ICP-RIE',\n",
       " 'process',\n",
       " 'Although',\n",
       " 'ICP',\n",
       " 'power',\n",
       " 'controls',\n",
       " 'plasma',\n",
       " 'density',\n",
       " 'additional',\n",
       " 'RF',\n",
       " 'source',\n",
       " 'coupled',\n",
       " 'capacitively',\n",
       " 'substrate',\n",
       " 'holder',\n",
       " 'adjust',\n",
       " 'substrate',\n",
       " 'bias',\n",
       " 'voltage',\n",
       " 'The',\n",
       " 'bias',\n",
       " 'voltage',\n",
       " 'substrate',\n",
       " 'generally',\n",
       " 'low',\n",
       " 'order',\n",
       " 'tens',\n",
       " 'volts',\n",
       " 'offering',\n",
       " 'low',\n",
       " 'velocity',\n",
       " 'ions',\n",
       " 'hit',\n",
       " 'substrate',\n",
       " 'As',\n",
       " 'result',\n",
       " 'ICP-RIE',\n",
       " 'method',\n",
       " 'provides',\n",
       " 'minimal',\n",
       " 'ion-induced',\n",
       " 'damage',\n",
       " 'Substrate',\n",
       " 'temperature',\n",
       " 'gas',\n",
       " 'flow',\n",
       " 'rate',\n",
       " 'chamber',\n",
       " 'pressure',\n",
       " 'kept',\n",
       " 'constant',\n",
       " 'values',\n",
       " '20',\n",
       " '°C',\n",
       " '0.6',\n",
       " 'Pa',\n",
       " '40',\n",
       " 'sccm',\n",
       " 'respectively',\n",
       " 'found',\n",
       " 'optimal',\n",
       " 'growth',\n",
       " 'ICP',\n",
       " 'RF',\n",
       " 'powers',\n",
       " 'varied',\n",
       " 'attain',\n",
       " 'desired',\n",
       " 'aspect',\n",
       " 'ratio',\n",
       " 'lowest',\n",
       " 'reflectance',\n",
       " 'Si',\n",
       " 'nanostructures',\n",
       " 'The',\n",
       " 'n-type',\n",
       " 'undoped',\n",
       " 'GaN-coated',\n",
       " 'sapphire',\n",
       " 'substrate',\n",
       " 'initially',\n",
       " 'cleaned',\n",
       " 'acetone',\n",
       " 'ethanol',\n",
       " 'deionized',\n",
       " 'water',\n",
       " 'respectively',\n",
       " 'blown',\n",
       " 'dry',\n",
       " 'N2',\n",
       " 'deposition',\n",
       " 'The',\n",
       " 'n-GaN',\n",
       " 'layer',\n",
       " 'thickness',\n",
       " '3.2',\n",
       " 'μm',\n",
       " 'resistivity',\n",
       " 'carrier',\n",
       " 'concentration',\n",
       " '9.13',\n",
       " '×',\n",
       " '10',\n",
       " '−2',\n",
       " 'Ω',\n",
       " 'cm',\n",
       " '1.78',\n",
       " '×',\n",
       " '10',\n",
       " '17',\n",
       " 'cm−3',\n",
       " 'respectively',\n",
       " 'Then',\n",
       " '80-nm',\n",
       " 'MZO',\n",
       " 'film',\n",
       " 'deposited',\n",
       " 'GaN-coated',\n",
       " 'sapphire',\n",
       " 'substrate',\n",
       " 'radio',\n",
       " 'frequency',\n",
       " 'reactive',\n",
       " 'magnetron',\n",
       " 'sputtering',\n",
       " '100',\n",
       " '°C',\n",
       " 'The',\n",
       " 'films',\n",
       " 'deposited',\n",
       " 'total',\n",
       " 'pressure',\n",
       " '2',\n",
       " 'Pa',\n",
       " 'power',\n",
       " 'sputtering',\n",
       " '120',\n",
       " 'W',\n",
       " 'Ar/O2',\n",
       " 'ratio',\n",
       " '08:03',\n",
       " 'Herein',\n",
       " 'thickness',\n",
       " 'MZO',\n",
       " 'film',\n",
       " 'measured',\n",
       " 'surface',\n",
       " 'profilometer',\n",
       " 'Talysurf',\n",
       " 'Serles',\n",
       " 'II',\n",
       " 'MZO',\n",
       " 'composition',\n",
       " 'determined',\n",
       " 'X-ray',\n",
       " 'photoelectron',\n",
       " 'spectroscopy',\n",
       " 'XPS',\n",
       " 'Before',\n",
       " 'deposition',\n",
       " 'target',\n",
       " 'presputtered',\n",
       " '10',\n",
       " 'min',\n",
       " 'remove',\n",
       " 'possible',\n",
       " 'contaminants',\n",
       " 'Then',\n",
       " '80-nm-thick',\n",
       " 'Au',\n",
       " 'electrodes',\n",
       " 'patterned',\n",
       " 'ZnO',\n",
       " 'surface',\n",
       " 'shadow',\n",
       " 'mask',\n",
       " 'radio',\n",
       " 'frequency',\n",
       " 'reactive',\n",
       " 'magnetron',\n",
       " 'sputtering',\n",
       " 'room',\n",
       " 'temperature',\n",
       " 'After',\n",
       " 'devices',\n",
       " 'annealed',\n",
       " '700',\n",
       " '°C',\n",
       " '2',\n",
       " 'h',\n",
       " 'air',\n",
       " 'Finally',\n",
       " 'get',\n",
       " 'large-area',\n",
       " 'Ohmic',\n",
       " 'contacts',\n",
       " 'surface',\n",
       " 'GaN',\n",
       " 'In',\n",
       " 'electrode',\n",
       " 'applied',\n",
       " 'The',\n",
       " 'Au',\n",
       " 'electrode',\n",
       " 'area',\n",
       " 'π',\n",
       " ...]"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "words = [str(word) for word in words_df.word.tolist()]\n",
    "words"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "#part-of-speech\n",
    "import chemdataextractor\n",
    "from chemdataextractor.nlp.pos import ChemCrfPosTagger\n",
    "\n",
    "cpt = ChemCrfPosTagger()\n",
    "words_df['POS'] = [tag[1] for tag in cpt.tag(words)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "from chemdataextractor.nlp.cem import CsDictCemTagger\n",
    "\n",
    "cct = CsDictCemTagger()\n",
    "words_df['CEM'] = [tag[1] for tag in cct.tag(words)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "from gensim.models import Word2Vec\n",
    "\n",
    "model = Word2Vec.load(\"word2vec_model/word2vec_embeddings-SNAPSHOT.model\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "171603"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(model.wv.vocab)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "        <script type=\"text/javascript\">\n",
       "        window.PlotlyConfig = {MathJaxConfig: 'local'};\n",
       "        if (window.MathJax) {MathJax.Hub.Config({SVG: {font: \"STIX-Web\"}});}\n",
       "        if (typeof require !== 'undefined') {\n",
       "        require.undef(\"plotly\");\n",
       "        requirejs.config({\n",
       "            paths: {\n",
       "                'plotly': ['https://cdn.plot.ly/plotly-latest.min']\n",
       "            }\n",
       "        });\n",
       "        require(['plotly'], function(Plotly) {\n",
       "            window._Plotly = Plotly;\n",
       "        });\n",
       "        }\n",
       "        </script>\n",
       "        "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.plotly.v1+json": {
       "config": {
        "linkText": "Export to plot.ly",
        "plotlyServerURL": "https://plot.ly",
        "showLink": false
       },
       "data": [
        {
         "mode": "text",
         "text": [
          "catechols",
          "eF",
          "-methylene",
          "gnidrager",
          "S-9",
          "S-8",
          "S-7",
          "S-6",
          "S-5",
          "S-4",
          "S-3",
          "Nampt",
          "S-1",
          "S-0",
          "27°",
          "epoxidation",
          "woods",
          "TBAP/acetonitrile",
          "PPh2",
          "stluserwere",
          "Electroplating",
          "dV/d(ln",
          "woody",
          "eB",
          "localized",
          "1R,2R)-N",
          "161.9",
          "161.8",
          "5988",
          "v(C0",
          "3-glycidoxypropyltrimethoxysilane",
          "N′-dimethylformamide",
          "161.1",
          "161.0",
          "161.3",
          "161.2",
          "161.5",
          "161.4",
          "161.7",
          "161.6",
          "1,800",
          "Cd2(2,2′-bpy)3]2",
          "Western",
          "copolyfluorenes",
          "crossbar",
          "imidate",
          "Me2NPY2",
          "12.2%",
          "Versuche",
          "Euro",
          "osteoblastic",
          "CuAlO2+x",
          "TiSc2N@C80",
          "1,803",
          "aspirator",
          "oea",
          "14MPa",
          "pages",
          "superantigenic",
          "aradius",
          "deniatboat",
          "S-2",
          "non",
          "replaces",
          "ketoimine",
          "hydrolyzate",
          "eT",
          "ciarbegla",
          "nanomotor",
          "v(CO",
          "v(CN",
          "Y12",
          "CdSe/CdTe",
          "methylaniline",
          "histomorphometry",
          "Kreuer",
          "wooden",
          "Pergamon",
          "stpmetta",
          ".ecnerefer",
          "II.EXPERIMENTALPROCEDURE",
          "Sack",
          "ZnO4",
          "TRAM",
          "circuitry",
          "nmr",
          "Pyrazine",
          "reliabilities",
          "G8/C28",
          "PQ4",
          "Transitions",
          "H0(R",
          "Pinkerton",
          "nonexponentiality",
          "27%",
          "gaskets",
          "C60-Pd",
          "0052",
          "dCOx2Amp",
          "hanging",
          "pillar[5]arene",
          "27-"
         ],
         "type": "scatter",
         "x": [
          28.845674514770508,
          -47.16980743408203,
          42.63121032714844,
          -29.887008666992188,
          -47.99568176269531,
          -39.40752410888672,
          -56.58651351928711,
          -63.99558639526367,
          -68.7457504272461,
          -75.58858489990234,
          -72.0445785522461,
          43.00864028930664,
          -31.76697540283203,
          18.743637084960938,
          10.803143501281738,
          -69.69586181640625,
          -58.25688934326172,
          35.99537658691406,
          -16.85546875,
          -7.642091274261475,
          66.98417663574219,
          4.73401403427124,
          -48.86732864379883,
          -33.483699798583984,
          16.129467010498047,
          18.72365379333496,
          -23.91835594177246,
          -28.111148834228516,
          16.75157356262207,
          50.14194869995117,
          31.055641174316406,
          -70.71699523925781,
          -18.262250900268555,
          -17.002283096313477,
          -26.46373176574707,
          -38.75592803955078,
          -39.67563247680664,
          -45.73739242553711,
          -29.627552032470703,
          -35.735233306884766,
          28.862829208374023,
          8.217620849609375,
          4.193431854248047,
          -9.813697814941406,
          69.5782699584961,
          49.587921142578125,
          8.720026969909668,
          -22.844345092773438,
          19.228729248046875,
          64.64379119873047,
          -36.691280364990234,
          8.658349990844727,
          5.891435623168945,
          -0.32137441635131836,
          39.06059646606445,
          -24.058609008789062,
          22.39830780029297,
          -9.826136589050293,
          16.79762077331543,
          38.81122970581055,
          -13.881563186645508,
          -81.21192932128906,
          -22.85445785522461,
          76.86507415771484,
          46.063392639160156,
          5.050357818603516,
          -36.126888275146484,
          -18.405132293701172,
          53.87387466430664,
          42.14859390258789,
          40.84473419189453,
          -10.777198791503906,
          5.861861705780029,
          57.61127853393555,
          76.0864486694336,
          -17.792064666748047,
          43.567081451416016,
          -2.8252596855163574,
          -24.341623306274414,
          -19.18968391418457,
          17.664859771728516,
          -8.303325653076172,
          -51.00422668457031,
          -5.017800331115723,
          81.12001037597656,
          62.82916259765625,
          14.739840507507324,
          16.285266876220703,
          -0.876698911190033,
          -1.9531183242797852,
          -39.124900817871094,
          13.315139770507812,
          29.970672607421875,
          28.072961807250977,
          -29.193098068237305,
          47.064388275146484,
          29.852497100830078,
          -1.716145396232605,
          3.494776725769043,
          52.27273941040039,
          27.0526123046875,
          -21.275888442993164
         ],
         "y": [
          -31.275108337402344,
          11.62476921081543,
          -8.659788131713867,
          29.365190505981445,
          -8.76157283782959,
          -16.050777435302734,
          -16.972957611083984,
          -7.690908908843994,
          -17.11359977722168,
          -9.162437438964844,
          -26.255023956298828,
          -35.1298828125,
          65.15489196777344,
          -46.472686767578125,
          86.98159790039062,
          21.319751739501953,
          65.79537200927734,
          15.009644508361816,
          -55.665611267089844,
          14.30746841430664,
          46.514190673828125,
          -12.711782455444336,
          61.790924072265625,
          5.5156569480896,
          68.7064437866211,
          -5.269098281860352,
          -109.10072326660156,
          -80.64579010009766,
          33.23601150512695,
          -68.41990661621094,
          -17.408994674682617,
          44.080135345458984,
          -97.70734405517578,
          -86.20098114013672,
          -90.96232604980469,
          -83.01917266845703,
          -103.76007843017578,
          -92.94261932373047,
          -100.74992370605469,
          -92.80380249023438,
          58.40885543823242,
          21.301170349121094,
          58.427467346191406,
          -11.28870677947998,
          24.587818145751953,
          -19.225637435913086,
          6.967172145843506,
          105.14604187011719,
          46.49256134033203,
          66.52420806884766,
          -38.7413215637207,
          -1.587644338607788,
          30.500991821289062,
          6.611763954162598,
          77.6138687133789,
          -16.410173416137695,
          14.672415733337402,
          69.41155242919922,
          22.914274215698242,
          27.1326961517334,
          21.84830665588379,
          -18.91303062438965,
          49.137332916259766,
          3.988431215286255,
          4.156271934509277,
          -57.91098403930664,
          15.687519073486328,
          31.20432472229004,
          17.74601936340332,
          -61.01316833496094,
          -69.0771255493164,
          3.254568099975586,
          43.40446472167969,
          -6.611498832702637,
          -17.15821075439453,
          -31.717975616455078,
          43.82760238647461,
          -23.887571334838867,
          20.745975494384766,
          12.272115707397461,
          5.208001613616943,
          48.01310348510742,
          30.986549377441406,
          34.16769790649414,
          26.84429931640625,
          -35.51292037963867,
          -29.163991928100586,
          -16.399324417114258,
          22.211889266967773,
          -3.3234963417053223,
          44.14533996582031,
          13.99830436706543,
          -4.301393032073975,
          6.383118629455566,
          100.63642883300781,
          55.923866271972656,
          37.53829574584961,
          -38.81023406982422,
          14.238512992858887,
          33.1667366027832,
          24.469831466674805,
          -2.659003257751465
         ]
        }
       ],
       "layout": {
        "template": {
         "data": {
          "bar": [
           {
            "error_x": {
             "color": "#2a3f5f"
            },
            "error_y": {
             "color": "#2a3f5f"
            },
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             }
            },
            "type": "bar"
           }
          ],
          "barpolar": [
           {
            "marker": {
             "line": {
              "color": "#E5ECF6",
              "width": 0.5
             }
            },
            "type": "barpolar"
           }
          ],
          "carpet": [
           {
            "aaxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "baxis": {
             "endlinecolor": "#2a3f5f",
             "gridcolor": "white",
             "linecolor": "white",
             "minorgridcolor": "white",
             "startlinecolor": "#2a3f5f"
            },
            "type": "carpet"
           }
          ],
          "choropleth": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "choropleth"
           }
          ],
          "contour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "contour"
           }
          ],
          "contourcarpet": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "contourcarpet"
           }
          ],
          "heatmap": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmap"
           }
          ],
          "heatmapgl": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "heatmapgl"
           }
          ],
          "histogram": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "histogram"
           }
          ],
          "histogram2d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2d"
           }
          ],
          "histogram2dcontour": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "histogram2dcontour"
           }
          ],
          "mesh3d": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "type": "mesh3d"
           }
          ],
          "parcoords": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "parcoords"
           }
          ],
          "pie": [
           {
            "automargin": true,
            "type": "pie"
           }
          ],
          "scatter": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter"
           }
          ],
          "scatter3d": [
           {
            "line": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatter3d"
           }
          ],
          "scattercarpet": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattercarpet"
           }
          ],
          "scattergeo": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergeo"
           }
          ],
          "scattergl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattergl"
           }
          ],
          "scattermapbox": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scattermapbox"
           }
          ],
          "scatterpolar": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolar"
           }
          ],
          "scatterpolargl": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterpolargl"
           }
          ],
          "scatterternary": [
           {
            "marker": {
             "colorbar": {
              "outlinewidth": 0,
              "ticks": ""
             }
            },
            "type": "scatterternary"
           }
          ],
          "surface": [
           {
            "colorbar": {
             "outlinewidth": 0,
             "ticks": ""
            },
            "colorscale": [
             [
              0,
              "#0d0887"
             ],
             [
              0.1111111111111111,
              "#46039f"
             ],
             [
              0.2222222222222222,
              "#7201a8"
             ],
             [
              0.3333333333333333,
              "#9c179e"
             ],
             [
              0.4444444444444444,
              "#bd3786"
             ],
             [
              0.5555555555555556,
              "#d8576b"
             ],
             [
              0.6666666666666666,
              "#ed7953"
             ],
             [
              0.7777777777777778,
              "#fb9f3a"
             ],
             [
              0.8888888888888888,
              "#fdca26"
             ],
             [
              1,
              "#f0f921"
             ]
            ],
            "type": "surface"
           }
          ],
          "table": [
           {
            "cells": {
             "fill": {
              "color": "#EBF0F8"
             },
             "line": {
              "color": "white"
             }
            },
            "header": {
             "fill": {
              "color": "#C8D4E3"
             },
             "line": {
              "color": "white"
             }
            },
            "type": "table"
           }
          ]
         },
         "layout": {
          "annotationdefaults": {
           "arrowcolor": "#2a3f5f",
           "arrowhead": 0,
           "arrowwidth": 1
          },
          "coloraxis": {
           "colorbar": {
            "outlinewidth": 0,
            "ticks": ""
           }
          },
          "colorscale": {
           "diverging": [
            [
             0,
             "#8e0152"
            ],
            [
             0.1,
             "#c51b7d"
            ],
            [
             0.2,
             "#de77ae"
            ],
            [
             0.3,
             "#f1b6da"
            ],
            [
             0.4,
             "#fde0ef"
            ],
            [
             0.5,
             "#f7f7f7"
            ],
            [
             0.6,
             "#e6f5d0"
            ],
            [
             0.7,
             "#b8e186"
            ],
            [
             0.8,
             "#7fbc41"
            ],
            [
             0.9,
             "#4d9221"
            ],
            [
             1,
             "#276419"
            ]
           ],
           "sequential": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ],
           "sequentialminus": [
            [
             0,
             "#0d0887"
            ],
            [
             0.1111111111111111,
             "#46039f"
            ],
            [
             0.2222222222222222,
             "#7201a8"
            ],
            [
             0.3333333333333333,
             "#9c179e"
            ],
            [
             0.4444444444444444,
             "#bd3786"
            ],
            [
             0.5555555555555556,
             "#d8576b"
            ],
            [
             0.6666666666666666,
             "#ed7953"
            ],
            [
             0.7777777777777778,
             "#fb9f3a"
            ],
            [
             0.8888888888888888,
             "#fdca26"
            ],
            [
             1,
             "#f0f921"
            ]
           ]
          },
          "colorway": [
           "#636efa",
           "#EF553B",
           "#00cc96",
           "#ab63fa",
           "#FFA15A",
           "#19d3f3",
           "#FF6692",
           "#B6E880",
           "#FF97FF",
           "#FECB52"
          ],
          "font": {
           "color": "#2a3f5f"
          },
          "geo": {
           "bgcolor": "white",
           "lakecolor": "white",
           "landcolor": "#E5ECF6",
           "showlakes": true,
           "showland": true,
           "subunitcolor": "white"
          },
          "hoverlabel": {
           "align": "left"
          },
          "hovermode": "closest",
          "mapbox": {
           "style": "light"
          },
          "paper_bgcolor": "white",
          "plot_bgcolor": "#E5ECF6",
          "polar": {
           "angularaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "radialaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "scene": {
           "xaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "yaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           },
           "zaxis": {
            "backgroundcolor": "#E5ECF6",
            "gridcolor": "white",
            "gridwidth": 2,
            "linecolor": "white",
            "showbackground": true,
            "ticks": "",
            "zerolinecolor": "white"
           }
          },
          "shapedefaults": {
           "line": {
            "color": "#2a3f5f"
           }
          },
          "ternary": {
           "aaxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "baxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           },
           "bgcolor": "#E5ECF6",
           "caxis": {
            "gridcolor": "white",
            "linecolor": "white",
            "ticks": ""
           }
          },
          "title": {
           "x": 0.05
          },
          "xaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          },
          "yaxis": {
           "automargin": true,
           "gridcolor": "white",
           "linecolor": "white",
           "ticks": "",
           "title": {
            "standoff": 15
           },
           "zerolinecolor": "white",
           "zerolinewidth": 2
          }
         }
        }
       }
      },
      "text/html": [
       "<div>\n",
       "        \n",
       "        \n",
       "            <div id=\"76b870d2-c11a-452b-ab94-d4b10e6ec46c\" class=\"plotly-graph-div\" style=\"height:525px; width:100%;\"></div>\n",
       "            <script type=\"text/javascript\">\n",
       "                require([\"plotly\"], function(Plotly) {\n",
       "                    window.PLOTLYENV=window.PLOTLYENV || {};\n",
       "                    \n",
       "                if (document.getElementById(\"76b870d2-c11a-452b-ab94-d4b10e6ec46c\")) {\n",
       "                    Plotly.newPlot(\n",
       "                        '76b870d2-c11a-452b-ab94-d4b10e6ec46c',\n",
       "                        [{\"mode\": \"text\", \"text\": [\"catechols\", \"eF\", \"-methylene\", \"gnidrager\", \"S-9\", \"S-8\", \"S-7\", \"S-6\", \"S-5\", \"S-4\", \"S-3\", \"Nampt\", \"S-1\", \"S-0\", \"27\\u00b0\", \"epoxidation\", \"woods\", \"TBAP/acetonitrile\", \"PPh2\", \"stluserwere\", \"Electroplating\", \"dV/d(ln\", \"woody\", \"eB\", \"localized\", \"1R,2R)-N\", \"161.9\", \"161.8\", \"5988\", \"v(C0\", \"3-glycidoxypropyltrimethoxysilane\", \"N\\u2032-dimethylformamide\", \"161.1\", \"161.0\", \"161.3\", \"161.2\", \"161.5\", \"161.4\", \"161.7\", \"161.6\", \"1,800\", \"Cd2(2,2\\u2032-bpy)3]2\", \"Western\", \"copolyfluorenes\", \"crossbar\", \"imidate\", \"Me2NPY2\", \"12.2%\", \"Versuche\", \"Euro\", \"osteoblastic\", \"CuAlO2+x\", \"TiSc2N@C80\", \"1,803\", \"aspirator\", \"oea\", \"14MPa\", \"pages\", \"superantigenic\", \"aradius\", \"deniatboat\", \"S-2\", \"non\", \"replaces\", \"ketoimine\", \"hydrolyzate\", \"eT\", \"ciarbegla\", \"nanomotor\", \"v(CO\", \"v(CN\", \"Y12\", \"CdSe/CdTe\", \"methylaniline\", \"histomorphometry\", \"Kreuer\", \"wooden\", \"Pergamon\", \"stpmetta\", \".ecnerefer\", \"II.EXPERIMENTALPROCEDURE\", \"Sack\", \"ZnO4\", \"TRAM\", \"circuitry\", \"nmr\", \"Pyrazine\", \"reliabilities\", \"G8/C28\", \"PQ4\", \"Transitions\", \"H0(R\", \"Pinkerton\", \"nonexponentiality\", \"27%\", \"gaskets\", \"C60-Pd\", \"0052\", \"dCOx2Amp\", \"hanging\", \"pillar[5]arene\", \"27-\"], \"type\": \"scatter\", \"x\": [28.845674514770508, -47.16980743408203, 42.63121032714844, -29.887008666992188, -47.99568176269531, -39.40752410888672, -56.58651351928711, -63.99558639526367, -68.7457504272461, -75.58858489990234, -72.0445785522461, 43.00864028930664, -31.76697540283203, 18.743637084960938, 10.803143501281738, -69.69586181640625, -58.25688934326172, 35.99537658691406, -16.85546875, -7.642091274261475, 66.98417663574219, 4.73401403427124, -48.86732864379883, -33.483699798583984, 16.129467010498047, 18.72365379333496, -23.91835594177246, -28.111148834228516, 16.75157356262207, 50.14194869995117, 31.055641174316406, -70.71699523925781, -18.262250900268555, -17.002283096313477, -26.46373176574707, -38.75592803955078, -39.67563247680664, -45.73739242553711, -29.627552032470703, -35.735233306884766, 28.862829208374023, 8.217620849609375, 4.193431854248047, -9.813697814941406, 69.5782699584961, 49.587921142578125, 8.720026969909668, -22.844345092773438, 19.228729248046875, 64.64379119873047, -36.691280364990234, 8.658349990844727, 5.891435623168945, -0.32137441635131836, 39.06059646606445, -24.058609008789062, 22.39830780029297, -9.826136589050293, 16.79762077331543, 38.81122970581055, -13.881563186645508, -81.21192932128906, -22.85445785522461, 76.86507415771484, 46.063392639160156, 5.050357818603516, -36.126888275146484, -18.405132293701172, 53.87387466430664, 42.14859390258789, 40.84473419189453, -10.777198791503906, 5.861861705780029, 57.61127853393555, 76.0864486694336, -17.792064666748047, 43.567081451416016, -2.8252596855163574, -24.341623306274414, -19.18968391418457, 17.664859771728516, -8.303325653076172, -51.00422668457031, -5.017800331115723, 81.12001037597656, 62.82916259765625, 14.739840507507324, 16.285266876220703, -0.876698911190033, -1.9531183242797852, -39.124900817871094, 13.315139770507812, 29.970672607421875, 28.072961807250977, -29.193098068237305, 47.064388275146484, 29.852497100830078, -1.716145396232605, 3.494776725769043, 52.27273941040039, 27.0526123046875, -21.275888442993164], \"y\": [-31.275108337402344, 11.62476921081543, -8.659788131713867, 29.365190505981445, -8.76157283782959, -16.050777435302734, -16.972957611083984, -7.690908908843994, -17.11359977722168, -9.162437438964844, -26.255023956298828, -35.1298828125, 65.15489196777344, -46.472686767578125, 86.98159790039062, 21.319751739501953, 65.79537200927734, 15.009644508361816, -55.665611267089844, 14.30746841430664, 46.514190673828125, -12.711782455444336, 61.790924072265625, 5.5156569480896, 68.7064437866211, -5.269098281860352, -109.10072326660156, -80.64579010009766, 33.23601150512695, -68.41990661621094, -17.408994674682617, 44.080135345458984, -97.70734405517578, -86.20098114013672, -90.96232604980469, -83.01917266845703, -103.76007843017578, -92.94261932373047, -100.74992370605469, -92.80380249023438, 58.40885543823242, 21.301170349121094, 58.427467346191406, -11.28870677947998, 24.587818145751953, -19.225637435913086, 6.967172145843506, 105.14604187011719, 46.49256134033203, 66.52420806884766, -38.7413215637207, -1.587644338607788, 30.500991821289062, 6.611763954162598, 77.6138687133789, -16.410173416137695, 14.672415733337402, 69.41155242919922, 22.914274215698242, 27.1326961517334, 21.84830665588379, -18.91303062438965, 49.137332916259766, 3.988431215286255, 4.156271934509277, -57.91098403930664, 15.687519073486328, 31.20432472229004, 17.74601936340332, -61.01316833496094, -69.0771255493164, 3.254568099975586, 43.40446472167969, -6.611498832702637, -17.15821075439453, -31.717975616455078, 43.82760238647461, -23.887571334838867, 20.745975494384766, 12.272115707397461, 5.208001613616943, 48.01310348510742, 30.986549377441406, 34.16769790649414, 26.84429931640625, -35.51292037963867, -29.163991928100586, -16.399324417114258, 22.211889266967773, -3.3234963417053223, 44.14533996582031, 13.99830436706543, -4.301393032073975, 6.383118629455566, 100.63642883300781, 55.923866271972656, 37.53829574584961, -38.81023406982422, 14.238512992858887, 33.1667366027832, 24.469831466674805, -2.659003257751465]}],\n",
       "                        {\"template\": {\"data\": {\"bar\": [{\"error_x\": {\"color\": \"#2a3f5f\"}, \"error_y\": {\"color\": \"#2a3f5f\"}, \"marker\": {\"line\": {\"color\": \"#E5ECF6\", \"width\": 0.5}}, \"type\": \"bar\"}], \"barpolar\": [{\"marker\": {\"line\": {\"color\": \"#E5ECF6\", \"width\": 0.5}}, \"type\": \"barpolar\"}], \"carpet\": [{\"aaxis\": {\"endlinecolor\": \"#2a3f5f\", \"gridcolor\": \"white\", \"linecolor\": \"white\", \"minorgridcolor\": \"white\", \"startlinecolor\": \"#2a3f5f\"}, \"baxis\": {\"endlinecolor\": \"#2a3f5f\", \"gridcolor\": \"white\", \"linecolor\": \"white\", \"minorgridcolor\": \"white\", \"startlinecolor\": \"#2a3f5f\"}, \"type\": \"carpet\"}], \"choropleth\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"type\": \"choropleth\"}], \"contour\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"contour\"}], \"contourcarpet\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"type\": \"contourcarpet\"}], \"heatmap\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"heatmap\"}], \"heatmapgl\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"heatmapgl\"}], \"histogram\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"histogram\"}], \"histogram2d\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"histogram2d\"}], \"histogram2dcontour\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"histogram2dcontour\"}], \"mesh3d\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"type\": \"mesh3d\"}], \"parcoords\": [{\"line\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"parcoords\"}], \"pie\": [{\"automargin\": true, \"type\": \"pie\"}], \"scatter\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scatter\"}], \"scatter3d\": [{\"line\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scatter3d\"}], \"scattercarpet\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scattercarpet\"}], \"scattergeo\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scattergeo\"}], \"scattergl\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scattergl\"}], \"scattermapbox\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scattermapbox\"}], \"scatterpolar\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scatterpolar\"}], \"scatterpolargl\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scatterpolargl\"}], \"scatterternary\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scatterternary\"}], \"surface\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"surface\"}], \"table\": [{\"cells\": {\"fill\": {\"color\": \"#EBF0F8\"}, \"line\": {\"color\": \"white\"}}, \"header\": {\"fill\": {\"color\": \"#C8D4E3\"}, \"line\": {\"color\": \"white\"}}, \"type\": \"table\"}]}, \"layout\": {\"annotationdefaults\": {\"arrowcolor\": \"#2a3f5f\", \"arrowhead\": 0, \"arrowwidth\": 1}, \"coloraxis\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"colorscale\": {\"diverging\": [[0, \"#8e0152\"], [0.1, \"#c51b7d\"], [0.2, \"#de77ae\"], [0.3, \"#f1b6da\"], [0.4, \"#fde0ef\"], [0.5, \"#f7f7f7\"], [0.6, \"#e6f5d0\"], [0.7, \"#b8e186\"], [0.8, \"#7fbc41\"], [0.9, \"#4d9221\"], [1, \"#276419\"]], \"sequential\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"sequentialminus\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]]}, \"colorway\": [\"#636efa\", \"#EF553B\", \"#00cc96\", \"#ab63fa\", \"#FFA15A\", \"#19d3f3\", \"#FF6692\", \"#B6E880\", \"#FF97FF\", \"#FECB52\"], \"font\": {\"color\": \"#2a3f5f\"}, \"geo\": {\"bgcolor\": \"white\", \"lakecolor\": \"white\", \"landcolor\": \"#E5ECF6\", \"showlakes\": true, \"showland\": true, \"subunitcolor\": \"white\"}, \"hoverlabel\": {\"align\": \"left\"}, \"hovermode\": \"closest\", \"mapbox\": {\"style\": \"light\"}, \"paper_bgcolor\": \"white\", \"plot_bgcolor\": \"#E5ECF6\", \"polar\": {\"angularaxis\": {\"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\"}, \"bgcolor\": \"#E5ECF6\", \"radialaxis\": {\"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\"}}, \"scene\": {\"xaxis\": {\"backgroundcolor\": \"#E5ECF6\", \"gridcolor\": \"white\", \"gridwidth\": 2, \"linecolor\": \"white\", \"showbackground\": true, \"ticks\": \"\", \"zerolinecolor\": \"white\"}, \"yaxis\": {\"backgroundcolor\": \"#E5ECF6\", \"gridcolor\": \"white\", \"gridwidth\": 2, \"linecolor\": \"white\", \"showbackground\": true, \"ticks\": \"\", \"zerolinecolor\": \"white\"}, \"zaxis\": {\"backgroundcolor\": \"#E5ECF6\", \"gridcolor\": \"white\", \"gridwidth\": 2, \"linecolor\": \"white\", \"showbackground\": true, \"ticks\": \"\", \"zerolinecolor\": \"white\"}}, \"shapedefaults\": {\"line\": {\"color\": \"#2a3f5f\"}}, \"ternary\": {\"aaxis\": {\"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\"}, \"baxis\": {\"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\"}, \"bgcolor\": \"#E5ECF6\", \"caxis\": {\"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\"}}, \"title\": {\"x\": 0.05}, \"xaxis\": {\"automargin\": true, \"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\", \"title\": {\"standoff\": 15}, \"zerolinecolor\": \"white\", \"zerolinewidth\": 2}, \"yaxis\": {\"automargin\": true, \"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\", \"title\": {\"standoff\": 15}, \"zerolinecolor\": \"white\", \"zerolinewidth\": 2}}}},\n",
       "                        {\"responsive\": true}\n",
       "                    ).then(function(){\n",
       "                            \n",
       "var gd = document.getElementById('76b870d2-c11a-452b-ab94-d4b10e6ec46c');\n",
       "var x = new MutationObserver(function (mutations, observer) {{\n",
       "        var display = window.getComputedStyle(gd).display;\n",
       "        if (!display || display === 'none') {{\n",
       "            console.log([gd, 'removed!']);\n",
       "            Plotly.purge(gd);\n",
       "            observer.disconnect();\n",
       "        }}\n",
       "}});\n",
       "\n",
       "// Listen for the removal of the full notebook cells\n",
       "var notebookContainer = gd.closest('#notebook-container');\n",
       "if (notebookContainer) {{\n",
       "    x.observe(notebookContainer, {childList: true});\n",
       "}}\n",
       "\n",
       "// Listen for the clearing of the current output cell\n",
       "var outputEl = gd.closest('.output');\n",
       "if (outputEl) {{\n",
       "    x.observe(outputEl, {childList: true});\n",
       "}}\n",
       "\n",
       "                        })\n",
       "                };\n",
       "                });\n",
       "            </script>\n",
       "        </div>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.decomposition import IncrementalPCA    # inital reduction\n",
    "from sklearn.manifold import TSNE                   # final reduction\n",
    "import numpy as np                                  # array handling\n",
    "\n",
    "\n",
    "def reduce_dimensions(model):\n",
    "    num_dimensions = 2  # final num dimensions (2D, 3D, etc)\n",
    "\n",
    "    vectors = [] # positions in vector space\n",
    "    labels = [] # keep track of words to label our data again later\n",
    "    for i, word in enumerate(model.wv.vocab):\n",
    "        vectors.append(model.wv[word])\n",
    "        labels.append(word)\n",
    "        if i>100:\n",
    "            break\n",
    "\n",
    "    # convert both lists into numpy vectors for reduction\n",
    "    vectors = np.asarray(vectors)\n",
    "    labels = np.asarray(labels)\n",
    "\n",
    "    # reduce using t-SNE\n",
    "    vectors = np.asarray(vectors)\n",
    "    tsne = TSNE(n_components=num_dimensions, random_state=0)\n",
    "    vectors = tsne.fit_transform(vectors)\n",
    "\n",
    "    x_vals = [v[0] for v in vectors]\n",
    "    y_vals = [v[1] for v in vectors]\n",
    "    return x_vals, y_vals, labels\n",
    "\n",
    "\n",
    "x_vals, y_vals, labels = reduce_dimensions(model)\n",
    "\n",
    "def plot_with_plotly(x_vals, y_vals, labels, plot_in_notebook=True):\n",
    "    import plotly\n",
    "    from plotly.offline import init_notebook_mode, iplot, plot\n",
    "    import plotly.graph_objs as go\n",
    "\n",
    "    trace = go.Scatter(x=x_vals, y=y_vals, mode='text', text=labels)\n",
    "    data = [trace]\n",
    "\n",
    "    if plot_in_notebook:\n",
    "        init_notebook_mode(connected=True)\n",
    "        iplot(data, filename='word-embedding-plot')\n",
    "    else:\n",
    "        plot(data, filename='word-embedding-plot.html')\n",
    "\n",
    "\n",
    "def plot_with_matplotlib(x_vals, y_vals, labels):\n",
    "    import matplotlib.pyplot as plt\n",
    "    import random\n",
    "\n",
    "    random.seed(0)\n",
    "\n",
    "    plt.figure(figsize=(12, 12))\n",
    "    plt.scatter(x_vals, y_vals)\n",
    "\n",
    "    #\n",
    "    # Label randomly subsampled 25 data points\n",
    "    #\n",
    "    indices = list(range(len(labels)))\n",
    "    selected_indices = random.sample(indices, 25)\n",
    "    for i in selected_indices:\n",
    "        plt.annotate(labels[i], (x_vals[i], y_vals[i]))\n",
    "\n",
    "try:\n",
    "    get_ipython()\n",
    "except Exception:\n",
    "    plot_function = plot_with_matplotlib\n",
    "else:\n",
    "    plot_function = plot_with_plotly\n",
    "\n",
    "plot_function(x_vals, y_vals, labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "word_vec_list = np.zeros([len(words), 100])\n",
    "notin = 0\n",
    "for i, word in enumerate(words):\n",
    "    try:\n",
    "        wordvec = model.wv.get_vector(word)\n",
    "        word_vec_list[i] = wordvec\n",
    "    except:\n",
    "        notin += 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "195"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "notin"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "words_df['isalpha'] = [word.isalpha() for word in words]\n",
    "words_df['isdigit'] = [word.isdigit() for word in words]\n",
    "words_df['isstop'] = [word.lower() in stop for word in words]\n",
    "words_df['charlen'] = [len(word) for word in words]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "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>word</th>\n",
       "      <th>label</th>\n",
       "      <th>POS</th>\n",
       "      <th>CEM</th>\n",
       "      <th>isalpha</th>\n",
       "      <th>isdigit</th>\n",
       "      <th>isstop</th>\n",
       "      <th>charlen</th>\n",
       "      <th>0wordvec</th>\n",
       "      <th>1wordvec</th>\n",
       "      <th>2wordvec</th>\n",
       "      <th>3wordvec</th>\n",
       "      <th>4wordvec</th>\n",
       "      <th>5wordvec</th>\n",
       "      <th>6wordvec</th>\n",
       "      <th>7wordvec</th>\n",
       "      <th>8wordvec</th>\n",
       "      <th>9wordvec</th>\n",
       "      <th>10wordvec</th>\n",
       "      <th>11wordvec</th>\n",
       "      <th>12wordvec</th>\n",
       "      <th>13wordvec</th>\n",
       "      <th>14wordvec</th>\n",
       "      <th>15wordvec</th>\n",
       "      <th>16wordvec</th>\n",
       "      <th>17wordvec</th>\n",
       "      <th>18wordvec</th>\n",
       "      <th>19wordvec</th>\n",
       "      <th>20wordvec</th>\n",
       "      <th>21wordvec</th>\n",
       "      <th>22wordvec</th>\n",
       "      <th>23wordvec</th>\n",
       "      <th>24wordvec</th>\n",
       "      <th>25wordvec</th>\n",
       "      <th>26wordvec</th>\n",
       "      <th>27wordvec</th>\n",
       "      <th>28wordvec</th>\n",
       "      <th>29wordvec</th>\n",
       "      <th>30wordvec</th>\n",
       "      <th>31wordvec</th>\n",
       "      <th>32wordvec</th>\n",
       "      <th>33wordvec</th>\n",
       "      <th>34wordvec</th>\n",
       "      <th>35wordvec</th>\n",
       "      <th>36wordvec</th>\n",
       "      <th>37wordvec</th>\n",
       "      <th>38wordvec</th>\n",
       "      <th>39wordvec</th>\n",
       "      <th>40wordvec</th>\n",
       "      <th>41wordvec</th>\n",
       "      <th>42wordvec</th>\n",
       "      <th>43wordvec</th>\n",
       "      <th>44wordvec</th>\n",
       "      <th>45wordvec</th>\n",
       "      <th>46wordvec</th>\n",
       "      <th>47wordvec</th>\n",
       "      <th>48wordvec</th>\n",
       "      <th>49wordvec</th>\n",
       "      <th>50wordvec</th>\n",
       "      <th>51wordvec</th>\n",
       "      <th>52wordvec</th>\n",
       "      <th>53wordvec</th>\n",
       "      <th>54wordvec</th>\n",
       "      <th>55wordvec</th>\n",
       "      <th>56wordvec</th>\n",
       "      <th>57wordvec</th>\n",
       "      <th>58wordvec</th>\n",
       "      <th>59wordvec</th>\n",
       "      <th>60wordvec</th>\n",
       "      <th>61wordvec</th>\n",
       "      <th>62wordvec</th>\n",
       "      <th>63wordvec</th>\n",
       "      <th>64wordvec</th>\n",
       "      <th>65wordvec</th>\n",
       "      <th>66wordvec</th>\n",
       "      <th>67wordvec</th>\n",
       "      <th>68wordvec</th>\n",
       "      <th>69wordvec</th>\n",
       "      <th>70wordvec</th>\n",
       "      <th>71wordvec</th>\n",
       "      <th>72wordvec</th>\n",
       "      <th>73wordvec</th>\n",
       "      <th>74wordvec</th>\n",
       "      <th>75wordvec</th>\n",
       "      <th>76wordvec</th>\n",
       "      <th>77wordvec</th>\n",
       "      <th>78wordvec</th>\n",
       "      <th>79wordvec</th>\n",
       "      <th>80wordvec</th>\n",
       "      <th>81wordvec</th>\n",
       "      <th>82wordvec</th>\n",
       "      <th>83wordvec</th>\n",
       "      <th>84wordvec</th>\n",
       "      <th>85wordvec</th>\n",
       "      <th>86wordvec</th>\n",
       "      <th>87wordvec</th>\n",
       "      <th>88wordvec</th>\n",
       "      <th>89wordvec</th>\n",
       "      <th>90wordvec</th>\n",
       "      <th>91wordvec</th>\n",
       "      <th>92wordvec</th>\n",
       "      <th>93wordvec</th>\n",
       "      <th>94wordvec</th>\n",
       "      <th>95wordvec</th>\n",
       "      <th>96wordvec</th>\n",
       "      <th>97wordvec</th>\n",
       "      <th>98wordvec</th>\n",
       "      <th>99wordvec</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Synthesis</td>\n",
       "      <td>7</td>\n",
       "      <td>NN</td>\n",
       "      <td>None</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>9</td>\n",
       "      <td>4.194923</td>\n",
       "      <td>-1.647826</td>\n",
       "      <td>0.285723</td>\n",
       "      <td>-0.403515</td>\n",
       "      <td>-1.439632</td>\n",
       "      <td>3.977208</td>\n",
       "      <td>0.003567</td>\n",
       "      <td>-0.402103</td>\n",
       "      <td>-3.030923</td>\n",
       "      <td>0.239334</td>\n",
       "      <td>1.903148</td>\n",
       "      <td>-0.066164</td>\n",
       "      <td>3.878762</td>\n",
       "      <td>-1.759037</td>\n",
       "      <td>-1.983543</td>\n",
       "      <td>0.494621</td>\n",
       "      <td>2.230356</td>\n",
       "      <td>-0.498870</td>\n",
       "      <td>-1.087159</td>\n",
       "      <td>3.058297</td>\n",
       "      <td>1.846964</td>\n",
       "      <td>-1.360210</td>\n",
       "      <td>1.749917</td>\n",
       "      <td>-2.791152</td>\n",
       "      <td>0.193650</td>\n",
       "      <td>-0.131373</td>\n",
       "      <td>3.656109</td>\n",
       "      <td>-2.829138</td>\n",
       "      <td>3.759925</td>\n",
       "      <td>-1.269079</td>\n",
       "      <td>0.627315</td>\n",
       "      <td>1.318691</td>\n",
       "      <td>0.717947</td>\n",
       "      <td>1.763228</td>\n",
       "      <td>-1.141046</td>\n",
       "      <td>-1.655796</td>\n",
       "      <td>1.603378</td>\n",
       "      <td>-0.041146</td>\n",
       "      <td>1.682122</td>\n",
       "      <td>0.533093</td>\n",
       "      <td>-2.147028</td>\n",
       "      <td>-3.696267</td>\n",
       "      <td>0.556257</td>\n",
       "      <td>-2.605197</td>\n",
       "      <td>-1.744037</td>\n",
       "      <td>-0.012125</td>\n",
       "      <td>1.727836</td>\n",
       "      <td>-2.956424</td>\n",
       "      <td>-2.908926</td>\n",
       "      <td>-0.137370</td>\n",
       "      <td>-1.366988</td>\n",
       "      <td>-0.372746</td>\n",
       "      <td>1.676232</td>\n",
       "      <td>-2.871509</td>\n",
       "      <td>-0.265448</td>\n",
       "      <td>-3.116419</td>\n",
       "      <td>-1.732071</td>\n",
       "      <td>1.233142</td>\n",
       "      <td>0.389019</td>\n",
       "      <td>-0.857282</td>\n",
       "      <td>0.655005</td>\n",
       "      <td>0.236637</td>\n",
       "      <td>-0.461650</td>\n",
       "      <td>-0.819598</td>\n",
       "      <td>2.007476</td>\n",
       "      <td>1.986614</td>\n",
       "      <td>0.888608</td>\n",
       "      <td>-0.255667</td>\n",
       "      <td>0.634797</td>\n",
       "      <td>-3.143937</td>\n",
       "      <td>-0.922885</td>\n",
       "      <td>1.092371</td>\n",
       "      <td>-0.503566</td>\n",
       "      <td>2.288535</td>\n",
       "      <td>0.165174</td>\n",
       "      <td>-0.450985</td>\n",
       "      <td>-2.547153</td>\n",
       "      <td>-1.913383</td>\n",
       "      <td>0.511201</td>\n",
       "      <td>0.009664</td>\n",
       "      <td>0.743061</td>\n",
       "      <td>1.960070</td>\n",
       "      <td>-0.286327</td>\n",
       "      <td>0.698199</td>\n",
       "      <td>1.170635</td>\n",
       "      <td>0.737539</td>\n",
       "      <td>-1.015091</td>\n",
       "      <td>0.795669</td>\n",
       "      <td>0.599087</td>\n",
       "      <td>0.084002</td>\n",
       "      <td>1.258723</td>\n",
       "      <td>-1.027135</td>\n",
       "      <td>-1.033529</td>\n",
       "      <td>-0.617139</td>\n",
       "      <td>1.604388</td>\n",
       "      <td>-1.584598</td>\n",
       "      <td>-0.309009</td>\n",
       "      <td>-0.456722</td>\n",
       "      <td>-0.142441</td>\n",
       "      <td>-1.925346</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Co−P</td>\n",
       "      <td>2</td>\n",
       "      <td>NN</td>\n",
       "      <td>None</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>4</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Nanocomposites</td>\n",
       "      <td>3</td>\n",
       "      <td>NNS</td>\n",
       "      <td>None</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>14</td>\n",
       "      <td>4.632976</td>\n",
       "      <td>-2.221284</td>\n",
       "      <td>2.954308</td>\n",
       "      <td>1.385346</td>\n",
       "      <td>0.276998</td>\n",
       "      <td>0.998676</td>\n",
       "      <td>-1.000307</td>\n",
       "      <td>-1.130656</td>\n",
       "      <td>-3.233622</td>\n",
       "      <td>-1.532178</td>\n",
       "      <td>1.014846</td>\n",
       "      <td>0.745873</td>\n",
       "      <td>0.497713</td>\n",
       "      <td>-0.525447</td>\n",
       "      <td>0.016174</td>\n",
       "      <td>1.495706</td>\n",
       "      <td>-2.816931</td>\n",
       "      <td>0.788923</td>\n",
       "      <td>1.062170</td>\n",
       "      <td>0.505834</td>\n",
       "      <td>1.476626</td>\n",
       "      <td>0.432519</td>\n",
       "      <td>0.142798</td>\n",
       "      <td>-3.296879</td>\n",
       "      <td>0.552678</td>\n",
       "      <td>3.410572</td>\n",
       "      <td>3.641243</td>\n",
       "      <td>-1.574522</td>\n",
       "      <td>2.891178</td>\n",
       "      <td>-2.327462</td>\n",
       "      <td>1.168447</td>\n",
       "      <td>1.355442</td>\n",
       "      <td>-1.172096</td>\n",
       "      <td>1.354260</td>\n",
       "      <td>-3.140076</td>\n",
       "      <td>1.723918</td>\n",
       "      <td>0.819414</td>\n",
       "      <td>-0.328178</td>\n",
       "      <td>1.247774</td>\n",
       "      <td>0.150336</td>\n",
       "      <td>-1.731371</td>\n",
       "      <td>-2.606534</td>\n",
       "      <td>-0.230035</td>\n",
       "      <td>-0.335319</td>\n",
       "      <td>0.599537</td>\n",
       "      <td>-0.109580</td>\n",
       "      <td>1.702124</td>\n",
       "      <td>-1.651153</td>\n",
       "      <td>-0.557687</td>\n",
       "      <td>-0.153378</td>\n",
       "      <td>0.125384</td>\n",
       "      <td>-0.683846</td>\n",
       "      <td>-1.033634</td>\n",
       "      <td>0.559948</td>\n",
       "      <td>-2.047792</td>\n",
       "      <td>0.659520</td>\n",
       "      <td>0.260303</td>\n",
       "      <td>-1.081487</td>\n",
       "      <td>0.531763</td>\n",
       "      <td>0.330068</td>\n",
       "      <td>0.171431</td>\n",
       "      <td>-1.254520</td>\n",
       "      <td>0.000297</td>\n",
       "      <td>-1.437393</td>\n",
       "      <td>0.821028</td>\n",
       "      <td>1.487649</td>\n",
       "      <td>1.457389</td>\n",
       "      <td>-3.244408</td>\n",
       "      <td>5.036873</td>\n",
       "      <td>-1.326810</td>\n",
       "      <td>0.022210</td>\n",
       "      <td>-0.656755</td>\n",
       "      <td>1.461387</td>\n",
       "      <td>-1.290582</td>\n",
       "      <td>-2.796442</td>\n",
       "      <td>-1.107028</td>\n",
       "      <td>-1.562787</td>\n",
       "      <td>-1.631077</td>\n",
       "      <td>-0.131813</td>\n",
       "      <td>-0.848237</td>\n",
       "      <td>-0.441429</td>\n",
       "      <td>2.140325</td>\n",
       "      <td>-1.728304</td>\n",
       "      <td>2.555290</td>\n",
       "      <td>-2.712211</td>\n",
       "      <td>-0.088766</td>\n",
       "      <td>-0.854533</td>\n",
       "      <td>-0.945716</td>\n",
       "      <td>0.874266</td>\n",
       "      <td>-1.725097</td>\n",
       "      <td>0.232114</td>\n",
       "      <td>-0.891776</td>\n",
       "      <td>0.654781</td>\n",
       "      <td>-0.291629</td>\n",
       "      <td>3.021340</td>\n",
       "      <td>-0.095630</td>\n",
       "      <td>1.525122</td>\n",
       "      <td>-0.445997</td>\n",
       "      <td>-0.487740</td>\n",
       "      <td>-1.329539</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Co−P</td>\n",
       "      <td>2</td>\n",
       "      <td>JJ</td>\n",
       "      <td>None</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>4</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>nanocomposites</td>\n",
       "      <td>3</td>\n",
       "      <td>NNS</td>\n",
       "      <td>None</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>14</td>\n",
       "      <td>3.984294</td>\n",
       "      <td>-2.487600</td>\n",
       "      <td>2.046803</td>\n",
       "      <td>-0.838005</td>\n",
       "      <td>-1.540583</td>\n",
       "      <td>1.617786</td>\n",
       "      <td>-0.550720</td>\n",
       "      <td>-2.355021</td>\n",
       "      <td>-1.709222</td>\n",
       "      <td>-0.899815</td>\n",
       "      <td>1.524484</td>\n",
       "      <td>0.645653</td>\n",
       "      <td>0.386420</td>\n",
       "      <td>-0.688321</td>\n",
       "      <td>-1.460670</td>\n",
       "      <td>0.596068</td>\n",
       "      <td>-0.853729</td>\n",
       "      <td>0.403138</td>\n",
       "      <td>0.899711</td>\n",
       "      <td>1.563543</td>\n",
       "      <td>0.230078</td>\n",
       "      <td>0.124216</td>\n",
       "      <td>0.741261</td>\n",
       "      <td>-2.846546</td>\n",
       "      <td>0.497974</td>\n",
       "      <td>3.247239</td>\n",
       "      <td>1.954929</td>\n",
       "      <td>1.075395</td>\n",
       "      <td>3.519975</td>\n",
       "      <td>-0.223869</td>\n",
       "      <td>-1.074260</td>\n",
       "      <td>-0.795728</td>\n",
       "      <td>-0.686169</td>\n",
       "      <td>0.152594</td>\n",
       "      <td>-2.866799</td>\n",
       "      <td>-0.041103</td>\n",
       "      <td>1.157811</td>\n",
       "      <td>0.479108</td>\n",
       "      <td>1.043930</td>\n",
       "      <td>1.422747</td>\n",
       "      <td>-0.621546</td>\n",
       "      <td>-1.455324</td>\n",
       "      <td>-0.130412</td>\n",
       "      <td>0.157529</td>\n",
       "      <td>1.881561</td>\n",
       "      <td>0.652456</td>\n",
       "      <td>0.471943</td>\n",
       "      <td>0.133345</td>\n",
       "      <td>-1.063320</td>\n",
       "      <td>1.245695</td>\n",
       "      <td>-1.612220</td>\n",
       "      <td>-0.627347</td>\n",
       "      <td>-0.866766</td>\n",
       "      <td>0.809690</td>\n",
       "      <td>-1.034540</td>\n",
       "      <td>0.373967</td>\n",
       "      <td>2.031811</td>\n",
       "      <td>-0.871307</td>\n",
       "      <td>0.176444</td>\n",
       "      <td>0.047214</td>\n",
       "      <td>0.921034</td>\n",
       "      <td>0.320219</td>\n",
       "      <td>1.137383</td>\n",
       "      <td>0.549271</td>\n",
       "      <td>1.489129</td>\n",
       "      <td>0.827529</td>\n",
       "      <td>0.568371</td>\n",
       "      <td>-1.896026</td>\n",
       "      <td>1.920248</td>\n",
       "      <td>-2.274963</td>\n",
       "      <td>0.183441</td>\n",
       "      <td>1.620190</td>\n",
       "      <td>0.300559</td>\n",
       "      <td>-0.311142</td>\n",
       "      <td>1.561803</td>\n",
       "      <td>1.458749</td>\n",
       "      <td>-2.109971</td>\n",
       "      <td>0.573890</td>\n",
       "      <td>0.272402</td>\n",
       "      <td>-1.493875</td>\n",
       "      <td>-1.155046</td>\n",
       "      <td>0.110574</td>\n",
       "      <td>-1.560961</td>\n",
       "      <td>4.151100</td>\n",
       "      <td>-1.825742</td>\n",
       "      <td>-0.382409</td>\n",
       "      <td>0.554005</td>\n",
       "      <td>-0.238090</td>\n",
       "      <td>1.643237</td>\n",
       "      <td>-0.887008</td>\n",
       "      <td>-0.552797</td>\n",
       "      <td>-0.682924</td>\n",
       "      <td>-2.940348</td>\n",
       "      <td>0.377087</td>\n",
       "      <td>3.398987</td>\n",
       "      <td>1.443565</td>\n",
       "      <td>1.623439</td>\n",
       "      <td>-1.769730</td>\n",
       "      <td>1.261726</td>\n",
       "      <td>0.829737</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             word  label  POS   CEM  isalpha  isdigit  isstop  charlen  \\\n",
       "0       Synthesis      7   NN  None     True    False   False        9   \n",
       "1            Co−P      2   NN  None    False    False   False        4   \n",
       "2  Nanocomposites      3  NNS  None     True    False   False       14   \n",
       "3            Co−P      2   JJ  None    False    False   False        4   \n",
       "4  nanocomposites      3  NNS  None     True    False   False       14   \n",
       "\n",
       "   0wordvec  1wordvec  2wordvec  3wordvec  4wordvec  5wordvec  6wordvec  \\\n",
       "0  4.194923 -1.647826  0.285723 -0.403515 -1.439632  3.977208  0.003567   \n",
       "1  0.000000  0.000000  0.000000  0.000000  0.000000  0.000000  0.000000   \n",
       "2  4.632976 -2.221284  2.954308  1.385346  0.276998  0.998676 -1.000307   \n",
       "3  0.000000  0.000000  0.000000  0.000000  0.000000  0.000000  0.000000   \n",
       "4  3.984294 -2.487600  2.046803 -0.838005 -1.540583  1.617786 -0.550720   \n",
       "\n",
       "   7wordvec  8wordvec  9wordvec  10wordvec  11wordvec  12wordvec  13wordvec  \\\n",
       "0 -0.402103 -3.030923  0.239334   1.903148  -0.066164   3.878762  -1.759037   \n",
       "1  0.000000  0.000000  0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2 -1.130656 -3.233622 -1.532178   1.014846   0.745873   0.497713  -0.525447   \n",
       "3  0.000000  0.000000  0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4 -2.355021 -1.709222 -0.899815   1.524484   0.645653   0.386420  -0.688321   \n",
       "\n",
       "   14wordvec  15wordvec  16wordvec  17wordvec  18wordvec  19wordvec  \\\n",
       "0  -1.983543   0.494621   2.230356  -0.498870  -1.087159   3.058297   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2   0.016174   1.495706  -2.816931   0.788923   1.062170   0.505834   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4  -1.460670   0.596068  -0.853729   0.403138   0.899711   1.563543   \n",
       "\n",
       "   20wordvec  21wordvec  22wordvec  23wordvec  24wordvec  25wordvec  \\\n",
       "0   1.846964  -1.360210   1.749917  -2.791152   0.193650  -0.131373   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2   1.476626   0.432519   0.142798  -3.296879   0.552678   3.410572   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4   0.230078   0.124216   0.741261  -2.846546   0.497974   3.247239   \n",
       "\n",
       "   26wordvec  27wordvec  28wordvec  29wordvec  30wordvec  31wordvec  \\\n",
       "0   3.656109  -2.829138   3.759925  -1.269079   0.627315   1.318691   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2   3.641243  -1.574522   2.891178  -2.327462   1.168447   1.355442   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4   1.954929   1.075395   3.519975  -0.223869  -1.074260  -0.795728   \n",
       "\n",
       "   32wordvec  33wordvec  34wordvec  35wordvec  36wordvec  37wordvec  \\\n",
       "0   0.717947   1.763228  -1.141046  -1.655796   1.603378  -0.041146   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2  -1.172096   1.354260  -3.140076   1.723918   0.819414  -0.328178   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4  -0.686169   0.152594  -2.866799  -0.041103   1.157811   0.479108   \n",
       "\n",
       "   38wordvec  39wordvec  40wordvec  41wordvec  42wordvec  43wordvec  \\\n",
       "0   1.682122   0.533093  -2.147028  -3.696267   0.556257  -2.605197   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2   1.247774   0.150336  -1.731371  -2.606534  -0.230035  -0.335319   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4   1.043930   1.422747  -0.621546  -1.455324  -0.130412   0.157529   \n",
       "\n",
       "   44wordvec  45wordvec  46wordvec  47wordvec  48wordvec  49wordvec  \\\n",
       "0  -1.744037  -0.012125   1.727836  -2.956424  -2.908926  -0.137370   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2   0.599537  -0.109580   1.702124  -1.651153  -0.557687  -0.153378   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4   1.881561   0.652456   0.471943   0.133345  -1.063320   1.245695   \n",
       "\n",
       "   50wordvec  51wordvec  52wordvec  53wordvec  54wordvec  55wordvec  \\\n",
       "0  -1.366988  -0.372746   1.676232  -2.871509  -0.265448  -3.116419   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2   0.125384  -0.683846  -1.033634   0.559948  -2.047792   0.659520   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4  -1.612220  -0.627347  -0.866766   0.809690  -1.034540   0.373967   \n",
       "\n",
       "   56wordvec  57wordvec  58wordvec  59wordvec  60wordvec  61wordvec  \\\n",
       "0  -1.732071   1.233142   0.389019  -0.857282   0.655005   0.236637   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2   0.260303  -1.081487   0.531763   0.330068   0.171431  -1.254520   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4   2.031811  -0.871307   0.176444   0.047214   0.921034   0.320219   \n",
       "\n",
       "   62wordvec  63wordvec  64wordvec  65wordvec  66wordvec  67wordvec  \\\n",
       "0  -0.461650  -0.819598   2.007476   1.986614   0.888608  -0.255667   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2   0.000297  -1.437393   0.821028   1.487649   1.457389  -3.244408   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4   1.137383   0.549271   1.489129   0.827529   0.568371  -1.896026   \n",
       "\n",
       "   68wordvec  69wordvec  70wordvec  71wordvec  72wordvec  73wordvec  \\\n",
       "0   0.634797  -3.143937  -0.922885   1.092371  -0.503566   2.288535   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2   5.036873  -1.326810   0.022210  -0.656755   1.461387  -1.290582   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4   1.920248  -2.274963   0.183441   1.620190   0.300559  -0.311142   \n",
       "\n",
       "   74wordvec  75wordvec  76wordvec  77wordvec  78wordvec  79wordvec  \\\n",
       "0   0.165174  -0.450985  -2.547153  -1.913383   0.511201   0.009664   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2  -2.796442  -1.107028  -1.562787  -1.631077  -0.131813  -0.848237   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4   1.561803   1.458749  -2.109971   0.573890   0.272402  -1.493875   \n",
       "\n",
       "   80wordvec  81wordvec  82wordvec  83wordvec  84wordvec  85wordvec  \\\n",
       "0   0.743061   1.960070  -0.286327   0.698199   1.170635   0.737539   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2  -0.441429   2.140325  -1.728304   2.555290  -2.712211  -0.088766   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4  -1.155046   0.110574  -1.560961   4.151100  -1.825742  -0.382409   \n",
       "\n",
       "   86wordvec  87wordvec  88wordvec  89wordvec  90wordvec  91wordvec  \\\n",
       "0  -1.015091   0.795669   0.599087   0.084002   1.258723  -1.027135   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2  -0.854533  -0.945716   0.874266  -1.725097   0.232114  -0.891776   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4   0.554005  -0.238090   1.643237  -0.887008  -0.552797  -0.682924   \n",
       "\n",
       "   92wordvec  93wordvec  94wordvec  95wordvec  96wordvec  97wordvec  \\\n",
       "0  -1.033529  -0.617139   1.604388  -1.584598  -0.309009  -0.456722   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2   0.654781  -0.291629   3.021340  -0.095630   1.525122  -0.445997   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4  -2.940348   0.377087   3.398987   1.443565   1.623439  -1.769730   \n",
       "\n",
       "   98wordvec  99wordvec  \n",
       "0  -0.142441  -1.925346  \n",
       "1   0.000000   0.000000  \n",
       "2  -0.487740  -1.329539  \n",
       "3   0.000000   0.000000  \n",
       "4   1.261726   0.829737  "
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "words_df2 = pd.concat([words_df.reset_index(drop=True), pd.DataFrame(word_vec_list).add_suffix('wordvec')], axis=1)\n",
    "words_df2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "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>word</th>\n",
       "      <th>label</th>\n",
       "      <th>POS</th>\n",
       "      <th>CEM</th>\n",
       "      <th>isalpha</th>\n",
       "      <th>isdigit</th>\n",
       "      <th>isstop</th>\n",
       "      <th>charlen</th>\n",
       "      <th>0wordvec</th>\n",
       "      <th>1wordvec</th>\n",
       "      <th>2wordvec</th>\n",
       "      <th>3wordvec</th>\n",
       "      <th>4wordvec</th>\n",
       "      <th>5wordvec</th>\n",
       "      <th>6wordvec</th>\n",
       "      <th>7wordvec</th>\n",
       "      <th>8wordvec</th>\n",
       "      <th>9wordvec</th>\n",
       "      <th>10wordvec</th>\n",
       "      <th>11wordvec</th>\n",
       "      <th>12wordvec</th>\n",
       "      <th>13wordvec</th>\n",
       "      <th>14wordvec</th>\n",
       "      <th>15wordvec</th>\n",
       "      <th>16wordvec</th>\n",
       "      <th>17wordvec</th>\n",
       "      <th>18wordvec</th>\n",
       "      <th>19wordvec</th>\n",
       "      <th>20wordvec</th>\n",
       "      <th>21wordvec</th>\n",
       "      <th>22wordvec</th>\n",
       "      <th>23wordvec</th>\n",
       "      <th>24wordvec</th>\n",
       "      <th>25wordvec</th>\n",
       "      <th>26wordvec</th>\n",
       "      <th>27wordvec</th>\n",
       "      <th>28wordvec</th>\n",
       "      <th>29wordvec</th>\n",
       "      <th>30wordvec</th>\n",
       "      <th>31wordvec</th>\n",
       "      <th>32wordvec</th>\n",
       "      <th>33wordvec</th>\n",
       "      <th>34wordvec</th>\n",
       "      <th>35wordvec</th>\n",
       "      <th>36wordvec</th>\n",
       "      <th>37wordvec</th>\n",
       "      <th>38wordvec</th>\n",
       "      <th>39wordvec</th>\n",
       "      <th>40wordvec</th>\n",
       "      <th>41wordvec</th>\n",
       "      <th>42wordvec</th>\n",
       "      <th>43wordvec</th>\n",
       "      <th>44wordvec</th>\n",
       "      <th>45wordvec</th>\n",
       "      <th>46wordvec</th>\n",
       "      <th>47wordvec</th>\n",
       "      <th>48wordvec</th>\n",
       "      <th>49wordvec</th>\n",
       "      <th>50wordvec</th>\n",
       "      <th>51wordvec</th>\n",
       "      <th>52wordvec</th>\n",
       "      <th>53wordvec</th>\n",
       "      <th>54wordvec</th>\n",
       "      <th>55wordvec</th>\n",
       "      <th>56wordvec</th>\n",
       "      <th>57wordvec</th>\n",
       "      <th>58wordvec</th>\n",
       "      <th>59wordvec</th>\n",
       "      <th>60wordvec</th>\n",
       "      <th>61wordvec</th>\n",
       "      <th>62wordvec</th>\n",
       "      <th>63wordvec</th>\n",
       "      <th>64wordvec</th>\n",
       "      <th>65wordvec</th>\n",
       "      <th>66wordvec</th>\n",
       "      <th>67wordvec</th>\n",
       "      <th>68wordvec</th>\n",
       "      <th>69wordvec</th>\n",
       "      <th>70wordvec</th>\n",
       "      <th>71wordvec</th>\n",
       "      <th>72wordvec</th>\n",
       "      <th>73wordvec</th>\n",
       "      <th>74wordvec</th>\n",
       "      <th>75wordvec</th>\n",
       "      <th>76wordvec</th>\n",
       "      <th>77wordvec</th>\n",
       "      <th>78wordvec</th>\n",
       "      <th>79wordvec</th>\n",
       "      <th>80wordvec</th>\n",
       "      <th>81wordvec</th>\n",
       "      <th>82wordvec</th>\n",
       "      <th>83wordvec</th>\n",
       "      <th>84wordvec</th>\n",
       "      <th>85wordvec</th>\n",
       "      <th>86wordvec</th>\n",
       "      <th>87wordvec</th>\n",
       "      <th>88wordvec</th>\n",
       "      <th>89wordvec</th>\n",
       "      <th>90wordvec</th>\n",
       "      <th>91wordvec</th>\n",
       "      <th>92wordvec</th>\n",
       "      <th>93wordvec</th>\n",
       "      <th>94wordvec</th>\n",
       "      <th>95wordvec</th>\n",
       "      <th>96wordvec</th>\n",
       "      <th>97wordvec</th>\n",
       "      <th>98wordvec</th>\n",
       "      <th>99wordvec</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Synthesis</td>\n",
       "      <td>7</td>\n",
       "      <td>9</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>9</td>\n",
       "      <td>4.194923</td>\n",
       "      <td>-1.647826</td>\n",
       "      <td>0.285723</td>\n",
       "      <td>-0.403515</td>\n",
       "      <td>-1.439632</td>\n",
       "      <td>3.977208</td>\n",
       "      <td>0.003567</td>\n",
       "      <td>-0.402103</td>\n",
       "      <td>-3.030923</td>\n",
       "      <td>0.239334</td>\n",
       "      <td>1.903148</td>\n",
       "      <td>-0.066164</td>\n",
       "      <td>3.878762</td>\n",
       "      <td>-1.759037</td>\n",
       "      <td>-1.983543</td>\n",
       "      <td>0.494621</td>\n",
       "      <td>2.230356</td>\n",
       "      <td>-0.498870</td>\n",
       "      <td>-1.087159</td>\n",
       "      <td>3.058297</td>\n",
       "      <td>1.846964</td>\n",
       "      <td>-1.360210</td>\n",
       "      <td>1.749917</td>\n",
       "      <td>-2.791152</td>\n",
       "      <td>0.193650</td>\n",
       "      <td>-0.131373</td>\n",
       "      <td>3.656109</td>\n",
       "      <td>-2.829138</td>\n",
       "      <td>3.759925</td>\n",
       "      <td>-1.269079</td>\n",
       "      <td>0.627315</td>\n",
       "      <td>1.318691</td>\n",
       "      <td>0.717947</td>\n",
       "      <td>1.763228</td>\n",
       "      <td>-1.141046</td>\n",
       "      <td>-1.655796</td>\n",
       "      <td>1.603378</td>\n",
       "      <td>-0.041146</td>\n",
       "      <td>1.682122</td>\n",
       "      <td>0.533093</td>\n",
       "      <td>-2.147028</td>\n",
       "      <td>-3.696267</td>\n",
       "      <td>0.556257</td>\n",
       "      <td>-2.605197</td>\n",
       "      <td>-1.744037</td>\n",
       "      <td>-0.012125</td>\n",
       "      <td>1.727836</td>\n",
       "      <td>-2.956424</td>\n",
       "      <td>-2.908926</td>\n",
       "      <td>-0.137370</td>\n",
       "      <td>-1.366988</td>\n",
       "      <td>-0.372746</td>\n",
       "      <td>1.676232</td>\n",
       "      <td>-2.871509</td>\n",
       "      <td>-0.265448</td>\n",
       "      <td>-3.116419</td>\n",
       "      <td>-1.732071</td>\n",
       "      <td>1.233142</td>\n",
       "      <td>0.389019</td>\n",
       "      <td>-0.857282</td>\n",
       "      <td>0.655005</td>\n",
       "      <td>0.236637</td>\n",
       "      <td>-0.461650</td>\n",
       "      <td>-0.819598</td>\n",
       "      <td>2.007476</td>\n",
       "      <td>1.986614</td>\n",
       "      <td>0.888608</td>\n",
       "      <td>-0.255667</td>\n",
       "      <td>0.634797</td>\n",
       "      <td>-3.143937</td>\n",
       "      <td>-0.922885</td>\n",
       "      <td>1.092371</td>\n",
       "      <td>-0.503566</td>\n",
       "      <td>2.288535</td>\n",
       "      <td>0.165174</td>\n",
       "      <td>-0.450985</td>\n",
       "      <td>-2.547153</td>\n",
       "      <td>-1.913383</td>\n",
       "      <td>0.511201</td>\n",
       "      <td>0.009664</td>\n",
       "      <td>0.743061</td>\n",
       "      <td>1.960070</td>\n",
       "      <td>-0.286327</td>\n",
       "      <td>0.698199</td>\n",
       "      <td>1.170635</td>\n",
       "      <td>0.737539</td>\n",
       "      <td>-1.015091</td>\n",
       "      <td>0.795669</td>\n",
       "      <td>0.599087</td>\n",
       "      <td>0.084002</td>\n",
       "      <td>1.258723</td>\n",
       "      <td>-1.027135</td>\n",
       "      <td>-1.033529</td>\n",
       "      <td>-0.617139</td>\n",
       "      <td>1.604388</td>\n",
       "      <td>-1.584598</td>\n",
       "      <td>-0.309009</td>\n",
       "      <td>-0.456722</td>\n",
       "      <td>-0.142441</td>\n",
       "      <td>-1.925346</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Co−P</td>\n",
       "      <td>2</td>\n",
       "      <td>9</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Nanocomposites</td>\n",
       "      <td>3</td>\n",
       "      <td>12</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>14</td>\n",
       "      <td>4.632976</td>\n",
       "      <td>-2.221284</td>\n",
       "      <td>2.954308</td>\n",
       "      <td>1.385346</td>\n",
       "      <td>0.276998</td>\n",
       "      <td>0.998676</td>\n",
       "      <td>-1.000307</td>\n",
       "      <td>-1.130656</td>\n",
       "      <td>-3.233622</td>\n",
       "      <td>-1.532178</td>\n",
       "      <td>1.014846</td>\n",
       "      <td>0.745873</td>\n",
       "      <td>0.497713</td>\n",
       "      <td>-0.525447</td>\n",
       "      <td>0.016174</td>\n",
       "      <td>1.495706</td>\n",
       "      <td>-2.816931</td>\n",
       "      <td>0.788923</td>\n",
       "      <td>1.062170</td>\n",
       "      <td>0.505834</td>\n",
       "      <td>1.476626</td>\n",
       "      <td>0.432519</td>\n",
       "      <td>0.142798</td>\n",
       "      <td>-3.296879</td>\n",
       "      <td>0.552678</td>\n",
       "      <td>3.410572</td>\n",
       "      <td>3.641243</td>\n",
       "      <td>-1.574522</td>\n",
       "      <td>2.891178</td>\n",
       "      <td>-2.327462</td>\n",
       "      <td>1.168447</td>\n",
       "      <td>1.355442</td>\n",
       "      <td>-1.172096</td>\n",
       "      <td>1.354260</td>\n",
       "      <td>-3.140076</td>\n",
       "      <td>1.723918</td>\n",
       "      <td>0.819414</td>\n",
       "      <td>-0.328178</td>\n",
       "      <td>1.247774</td>\n",
       "      <td>0.150336</td>\n",
       "      <td>-1.731371</td>\n",
       "      <td>-2.606534</td>\n",
       "      <td>-0.230035</td>\n",
       "      <td>-0.335319</td>\n",
       "      <td>0.599537</td>\n",
       "      <td>-0.109580</td>\n",
       "      <td>1.702124</td>\n",
       "      <td>-1.651153</td>\n",
       "      <td>-0.557687</td>\n",
       "      <td>-0.153378</td>\n",
       "      <td>0.125384</td>\n",
       "      <td>-0.683846</td>\n",
       "      <td>-1.033634</td>\n",
       "      <td>0.559948</td>\n",
       "      <td>-2.047792</td>\n",
       "      <td>0.659520</td>\n",
       "      <td>0.260303</td>\n",
       "      <td>-1.081487</td>\n",
       "      <td>0.531763</td>\n",
       "      <td>0.330068</td>\n",
       "      <td>0.171431</td>\n",
       "      <td>-1.254520</td>\n",
       "      <td>0.000297</td>\n",
       "      <td>-1.437393</td>\n",
       "      <td>0.821028</td>\n",
       "      <td>1.487649</td>\n",
       "      <td>1.457389</td>\n",
       "      <td>-3.244408</td>\n",
       "      <td>5.036873</td>\n",
       "      <td>-1.326810</td>\n",
       "      <td>0.022210</td>\n",
       "      <td>-0.656755</td>\n",
       "      <td>1.461387</td>\n",
       "      <td>-1.290582</td>\n",
       "      <td>-2.796442</td>\n",
       "      <td>-1.107028</td>\n",
       "      <td>-1.562787</td>\n",
       "      <td>-1.631077</td>\n",
       "      <td>-0.131813</td>\n",
       "      <td>-0.848237</td>\n",
       "      <td>-0.441429</td>\n",
       "      <td>2.140325</td>\n",
       "      <td>-1.728304</td>\n",
       "      <td>2.555290</td>\n",
       "      <td>-2.712211</td>\n",
       "      <td>-0.088766</td>\n",
       "      <td>-0.854533</td>\n",
       "      <td>-0.945716</td>\n",
       "      <td>0.874266</td>\n",
       "      <td>-1.725097</td>\n",
       "      <td>0.232114</td>\n",
       "      <td>-0.891776</td>\n",
       "      <td>0.654781</td>\n",
       "      <td>-0.291629</td>\n",
       "      <td>3.021340</td>\n",
       "      <td>-0.095630</td>\n",
       "      <td>1.525122</td>\n",
       "      <td>-0.445997</td>\n",
       "      <td>-0.487740</td>\n",
       "      <td>-1.329539</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Co−P</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>-1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>nanocomposites</td>\n",
       "      <td>3</td>\n",
       "      <td>12</td>\n",
       "      <td>-1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>14</td>\n",
       "      <td>3.984294</td>\n",
       "      <td>-2.487600</td>\n",
       "      <td>2.046803</td>\n",
       "      <td>-0.838005</td>\n",
       "      <td>-1.540583</td>\n",
       "      <td>1.617786</td>\n",
       "      <td>-0.550720</td>\n",
       "      <td>-2.355021</td>\n",
       "      <td>-1.709222</td>\n",
       "      <td>-0.899815</td>\n",
       "      <td>1.524484</td>\n",
       "      <td>0.645653</td>\n",
       "      <td>0.386420</td>\n",
       "      <td>-0.688321</td>\n",
       "      <td>-1.460670</td>\n",
       "      <td>0.596068</td>\n",
       "      <td>-0.853729</td>\n",
       "      <td>0.403138</td>\n",
       "      <td>0.899711</td>\n",
       "      <td>1.563543</td>\n",
       "      <td>0.230078</td>\n",
       "      <td>0.124216</td>\n",
       "      <td>0.741261</td>\n",
       "      <td>-2.846546</td>\n",
       "      <td>0.497974</td>\n",
       "      <td>3.247239</td>\n",
       "      <td>1.954929</td>\n",
       "      <td>1.075395</td>\n",
       "      <td>3.519975</td>\n",
       "      <td>-0.223869</td>\n",
       "      <td>-1.074260</td>\n",
       "      <td>-0.795728</td>\n",
       "      <td>-0.686169</td>\n",
       "      <td>0.152594</td>\n",
       "      <td>-2.866799</td>\n",
       "      <td>-0.041103</td>\n",
       "      <td>1.157811</td>\n",
       "      <td>0.479108</td>\n",
       "      <td>1.043930</td>\n",
       "      <td>1.422747</td>\n",
       "      <td>-0.621546</td>\n",
       "      <td>-1.455324</td>\n",
       "      <td>-0.130412</td>\n",
       "      <td>0.157529</td>\n",
       "      <td>1.881561</td>\n",
       "      <td>0.652456</td>\n",
       "      <td>0.471943</td>\n",
       "      <td>0.133345</td>\n",
       "      <td>-1.063320</td>\n",
       "      <td>1.245695</td>\n",
       "      <td>-1.612220</td>\n",
       "      <td>-0.627347</td>\n",
       "      <td>-0.866766</td>\n",
       "      <td>0.809690</td>\n",
       "      <td>-1.034540</td>\n",
       "      <td>0.373967</td>\n",
       "      <td>2.031811</td>\n",
       "      <td>-0.871307</td>\n",
       "      <td>0.176444</td>\n",
       "      <td>0.047214</td>\n",
       "      <td>0.921034</td>\n",
       "      <td>0.320219</td>\n",
       "      <td>1.137383</td>\n",
       "      <td>0.549271</td>\n",
       "      <td>1.489129</td>\n",
       "      <td>0.827529</td>\n",
       "      <td>0.568371</td>\n",
       "      <td>-1.896026</td>\n",
       "      <td>1.920248</td>\n",
       "      <td>-2.274963</td>\n",
       "      <td>0.183441</td>\n",
       "      <td>1.620190</td>\n",
       "      <td>0.300559</td>\n",
       "      <td>-0.311142</td>\n",
       "      <td>1.561803</td>\n",
       "      <td>1.458749</td>\n",
       "      <td>-2.109971</td>\n",
       "      <td>0.573890</td>\n",
       "      <td>0.272402</td>\n",
       "      <td>-1.493875</td>\n",
       "      <td>-1.155046</td>\n",
       "      <td>0.110574</td>\n",
       "      <td>-1.560961</td>\n",
       "      <td>4.151100</td>\n",
       "      <td>-1.825742</td>\n",
       "      <td>-0.382409</td>\n",
       "      <td>0.554005</td>\n",
       "      <td>-0.238090</td>\n",
       "      <td>1.643237</td>\n",
       "      <td>-0.887008</td>\n",
       "      <td>-0.552797</td>\n",
       "      <td>-0.682924</td>\n",
       "      <td>-2.940348</td>\n",
       "      <td>0.377087</td>\n",
       "      <td>3.398987</td>\n",
       "      <td>1.443565</td>\n",
       "      <td>1.623439</td>\n",
       "      <td>-1.769730</td>\n",
       "      <td>1.261726</td>\n",
       "      <td>0.829737</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             word  label  POS  CEM  isalpha  isdigit  isstop  charlen  \\\n",
       "0       Synthesis      7    9   -1        1        0       0        9   \n",
       "1            Co−P      2    9   -1        0        0       0        4   \n",
       "2  Nanocomposites      3   12   -1        1        0       0       14   \n",
       "3            Co−P      2    5   -1        0        0       0        4   \n",
       "4  nanocomposites      3   12   -1        1        0       0       14   \n",
       "\n",
       "   0wordvec  1wordvec  2wordvec  3wordvec  4wordvec  5wordvec  6wordvec  \\\n",
       "0  4.194923 -1.647826  0.285723 -0.403515 -1.439632  3.977208  0.003567   \n",
       "1  0.000000  0.000000  0.000000  0.000000  0.000000  0.000000  0.000000   \n",
       "2  4.632976 -2.221284  2.954308  1.385346  0.276998  0.998676 -1.000307   \n",
       "3  0.000000  0.000000  0.000000  0.000000  0.000000  0.000000  0.000000   \n",
       "4  3.984294 -2.487600  2.046803 -0.838005 -1.540583  1.617786 -0.550720   \n",
       "\n",
       "   7wordvec  8wordvec  9wordvec  10wordvec  11wordvec  12wordvec  13wordvec  \\\n",
       "0 -0.402103 -3.030923  0.239334   1.903148  -0.066164   3.878762  -1.759037   \n",
       "1  0.000000  0.000000  0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2 -1.130656 -3.233622 -1.532178   1.014846   0.745873   0.497713  -0.525447   \n",
       "3  0.000000  0.000000  0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4 -2.355021 -1.709222 -0.899815   1.524484   0.645653   0.386420  -0.688321   \n",
       "\n",
       "   14wordvec  15wordvec  16wordvec  17wordvec  18wordvec  19wordvec  \\\n",
       "0  -1.983543   0.494621   2.230356  -0.498870  -1.087159   3.058297   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2   0.016174   1.495706  -2.816931   0.788923   1.062170   0.505834   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4  -1.460670   0.596068  -0.853729   0.403138   0.899711   1.563543   \n",
       "\n",
       "   20wordvec  21wordvec  22wordvec  23wordvec  24wordvec  25wordvec  \\\n",
       "0   1.846964  -1.360210   1.749917  -2.791152   0.193650  -0.131373   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2   1.476626   0.432519   0.142798  -3.296879   0.552678   3.410572   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4   0.230078   0.124216   0.741261  -2.846546   0.497974   3.247239   \n",
       "\n",
       "   26wordvec  27wordvec  28wordvec  29wordvec  30wordvec  31wordvec  \\\n",
       "0   3.656109  -2.829138   3.759925  -1.269079   0.627315   1.318691   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2   3.641243  -1.574522   2.891178  -2.327462   1.168447   1.355442   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4   1.954929   1.075395   3.519975  -0.223869  -1.074260  -0.795728   \n",
       "\n",
       "   32wordvec  33wordvec  34wordvec  35wordvec  36wordvec  37wordvec  \\\n",
       "0   0.717947   1.763228  -1.141046  -1.655796   1.603378  -0.041146   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2  -1.172096   1.354260  -3.140076   1.723918   0.819414  -0.328178   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4  -0.686169   0.152594  -2.866799  -0.041103   1.157811   0.479108   \n",
       "\n",
       "   38wordvec  39wordvec  40wordvec  41wordvec  42wordvec  43wordvec  \\\n",
       "0   1.682122   0.533093  -2.147028  -3.696267   0.556257  -2.605197   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2   1.247774   0.150336  -1.731371  -2.606534  -0.230035  -0.335319   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4   1.043930   1.422747  -0.621546  -1.455324  -0.130412   0.157529   \n",
       "\n",
       "   44wordvec  45wordvec  46wordvec  47wordvec  48wordvec  49wordvec  \\\n",
       "0  -1.744037  -0.012125   1.727836  -2.956424  -2.908926  -0.137370   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2   0.599537  -0.109580   1.702124  -1.651153  -0.557687  -0.153378   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4   1.881561   0.652456   0.471943   0.133345  -1.063320   1.245695   \n",
       "\n",
       "   50wordvec  51wordvec  52wordvec  53wordvec  54wordvec  55wordvec  \\\n",
       "0  -1.366988  -0.372746   1.676232  -2.871509  -0.265448  -3.116419   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2   0.125384  -0.683846  -1.033634   0.559948  -2.047792   0.659520   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4  -1.612220  -0.627347  -0.866766   0.809690  -1.034540   0.373967   \n",
       "\n",
       "   56wordvec  57wordvec  58wordvec  59wordvec  60wordvec  61wordvec  \\\n",
       "0  -1.732071   1.233142   0.389019  -0.857282   0.655005   0.236637   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2   0.260303  -1.081487   0.531763   0.330068   0.171431  -1.254520   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4   2.031811  -0.871307   0.176444   0.047214   0.921034   0.320219   \n",
       "\n",
       "   62wordvec  63wordvec  64wordvec  65wordvec  66wordvec  67wordvec  \\\n",
       "0  -0.461650  -0.819598   2.007476   1.986614   0.888608  -0.255667   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2   0.000297  -1.437393   0.821028   1.487649   1.457389  -3.244408   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4   1.137383   0.549271   1.489129   0.827529   0.568371  -1.896026   \n",
       "\n",
       "   68wordvec  69wordvec  70wordvec  71wordvec  72wordvec  73wordvec  \\\n",
       "0   0.634797  -3.143937  -0.922885   1.092371  -0.503566   2.288535   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2   5.036873  -1.326810   0.022210  -0.656755   1.461387  -1.290582   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4   1.920248  -2.274963   0.183441   1.620190   0.300559  -0.311142   \n",
       "\n",
       "   74wordvec  75wordvec  76wordvec  77wordvec  78wordvec  79wordvec  \\\n",
       "0   0.165174  -0.450985  -2.547153  -1.913383   0.511201   0.009664   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2  -2.796442  -1.107028  -1.562787  -1.631077  -0.131813  -0.848237   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4   1.561803   1.458749  -2.109971   0.573890   0.272402  -1.493875   \n",
       "\n",
       "   80wordvec  81wordvec  82wordvec  83wordvec  84wordvec  85wordvec  \\\n",
       "0   0.743061   1.960070  -0.286327   0.698199   1.170635   0.737539   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2  -0.441429   2.140325  -1.728304   2.555290  -2.712211  -0.088766   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4  -1.155046   0.110574  -1.560961   4.151100  -1.825742  -0.382409   \n",
       "\n",
       "   86wordvec  87wordvec  88wordvec  89wordvec  90wordvec  91wordvec  \\\n",
       "0  -1.015091   0.795669   0.599087   0.084002   1.258723  -1.027135   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2  -0.854533  -0.945716   0.874266  -1.725097   0.232114  -0.891776   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4   0.554005  -0.238090   1.643237  -0.887008  -0.552797  -0.682924   \n",
       "\n",
       "   92wordvec  93wordvec  94wordvec  95wordvec  96wordvec  97wordvec  \\\n",
       "0  -1.033529  -0.617139   1.604388  -1.584598  -0.309009  -0.456722   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2   0.654781  -0.291629   3.021340  -0.095630   1.525122  -0.445997   \n",
       "3   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "4  -2.940348   0.377087   3.398987   1.443565   1.623439  -1.769730   \n",
       "\n",
       "   98wordvec  99wordvec  \n",
       "0  -0.142441  -1.925346  \n",
       "1   0.000000   0.000000  \n",
       "2  -0.487740  -1.329539  \n",
       "3   0.000000   0.000000  \n",
       "4   1.261726   0.829737  "
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "for col in words_df2.columns[2:7]:\n",
    "    words_df2[col] = words_df2[col].astype('category').cat.codes\n",
    "words_df2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "words_df2['CEM'] = words_df2['CEM'] + 1\n",
    "words_df2.drop('word', axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import accuracy_score, f1_score, roc_auc_score, roc_curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = words_df2.drop('label', axis=1)\n",
    "Y = words_df2['label']\n",
    "trainX, testX, trainY, testY = train_test_split(X, Y, test_size=0.2, random_state=96)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/ynyuan/anaconda3/envs/nlp/lib/python3.6/site-packages/sklearn/svm/base.py:193: FutureWarning:\n",
      "\n",
      "The default value of gamma will change from 'auto' to 'scale' in version 0.22 to account better for unscaled features. Set gamma explicitly to 'auto' or 'scale' to avoid this warning.\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "SVC(C=0.1, cache_size=200, class_weight=None, coef0=0.0,\n",
       "    decision_function_shape='ovr', degree=3, gamma='auto_deprecated',\n",
       "    kernel='rbf', max_iter=-1, probability=True, random_state=120,\n",
       "    shrinking=True, tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "reg = SVC(C=0.1, probability=True, random_state=120)\n",
    "reg.fit(trainX, trainY)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.6706586826347305\n",
      "0.6711509715994021\n",
      "0.977190956317156\n",
      "0.9584630177041045\n"
     ]
    }
   ],
   "source": [
    "print(reg.score(trainX, trainY))\n",
    "print(reg.score(testX, testY))\n",
    "print(roc_auc_score(pd.get_dummies(trainY, prefix='label'), reg.predict_proba(trainX), average='micro'))\n",
    "print(roc_auc_score(pd.get_dummies(testY, prefix='label'), reg.predict_proba(testX), average='micro'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',\n",
       "                       max_depth=7, max_features='auto', max_leaf_nodes=None,\n",
       "                       min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                       min_samples_leaf=1, min_samples_split=2,\n",
       "                       min_weight_fraction_leaf=0.0, n_estimators=100,\n",
       "                       n_jobs=None, oob_score=False, random_state=10, verbose=0,\n",
       "                       warm_start=False)"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf = RandomForestClassifier(n_estimators=100, max_depth=7, random_state=10)\n",
    "rf.fit(trainX, trainY)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7657185628742516\n",
      "0.7399103139013453\n",
      "0.9861056213732715\n",
      "0.9693770793254999\n"
     ]
    }
   ],
   "source": [
    "print(accuracy_score(trainY, rf.predict(trainX)))\n",
    "print(accuracy_score(testY, rf.predict(testX)))\n",
    "print(roc_auc_score(pd.get_dummies(trainY, prefix='label'), rf.predict_proba(trainX), average='micro'))\n",
    "print(roc_auc_score(pd.get_dummies(testY, prefix='label'), rf.predict_proba(testX), average='micro'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/ynyuan/anaconda3/envs/nlp/lib/python3.6/site-packages/lightgbm/engine.py:153: UserWarning: Found `early_stopping_round` in params. Will use it instead of argument\n",
      "  warnings.warn(\"Found `{}` in params. Will use it instead of argument\".format(alias))\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1]\tTrain's multi_logloss: 1.71399\tEvaluation's multi_logloss: 1.74383\n",
      "Training until validation scores don't improve for 10 rounds\n",
      "[2]\tTrain's multi_logloss: 1.67098\tEvaluation's multi_logloss: 1.70292\n",
      "[3]\tTrain's multi_logloss: 1.63181\tEvaluation's multi_logloss: 1.66634\n",
      "[4]\tTrain's multi_logloss: 1.59481\tEvaluation's multi_logloss: 1.63209\n",
      "[5]\tTrain's multi_logloss: 1.56007\tEvaluation's multi_logloss: 1.60122\n",
      "[6]\tTrain's multi_logloss: 1.52735\tEvaluation's multi_logloss: 1.57172\n",
      "[7]\tTrain's multi_logloss: 1.49612\tEvaluation's multi_logloss: 1.54446\n",
      "[8]\tTrain's multi_logloss: 1.4669\tEvaluation's multi_logloss: 1.51827\n",
      "[9]\tTrain's multi_logloss: 1.43902\tEvaluation's multi_logloss: 1.4938\n",
      "[10]\tTrain's multi_logloss: 1.41253\tEvaluation's multi_logloss: 1.47062\n",
      "[11]\tTrain's multi_logloss: 1.3874\tEvaluation's multi_logloss: 1.44867\n",
      "[12]\tTrain's multi_logloss: 1.36287\tEvaluation's multi_logloss: 1.42781\n",
      "[13]\tTrain's multi_logloss: 1.33988\tEvaluation's multi_logloss: 1.40786\n",
      "[14]\tTrain's multi_logloss: 1.31754\tEvaluation's multi_logloss: 1.38936\n",
      "[15]\tTrain's multi_logloss: 1.29656\tEvaluation's multi_logloss: 1.37141\n",
      "[16]\tTrain's multi_logloss: 1.27577\tEvaluation's multi_logloss: 1.35421\n",
      "[17]\tTrain's multi_logloss: 1.25596\tEvaluation's multi_logloss: 1.33725\n",
      "[18]\tTrain's multi_logloss: 1.23672\tEvaluation's multi_logloss: 1.32158\n",
      "[19]\tTrain's multi_logloss: 1.21781\tEvaluation's multi_logloss: 1.30587\n",
      "[20]\tTrain's multi_logloss: 1.19996\tEvaluation's multi_logloss: 1.29072\n",
      "[21]\tTrain's multi_logloss: 1.18248\tEvaluation's multi_logloss: 1.27645\n",
      "[22]\tTrain's multi_logloss: 1.16541\tEvaluation's multi_logloss: 1.26247\n",
      "[23]\tTrain's multi_logloss: 1.14887\tEvaluation's multi_logloss: 1.24922\n",
      "[24]\tTrain's multi_logloss: 1.13297\tEvaluation's multi_logloss: 1.23625\n",
      "[25]\tTrain's multi_logloss: 1.1177\tEvaluation's multi_logloss: 1.22391\n",
      "[26]\tTrain's multi_logloss: 1.10269\tEvaluation's multi_logloss: 1.21187\n",
      "[27]\tTrain's multi_logloss: 1.0883\tEvaluation's multi_logloss: 1.20052\n",
      "[28]\tTrain's multi_logloss: 1.07449\tEvaluation's multi_logloss: 1.18903\n",
      "[29]\tTrain's multi_logloss: 1.06102\tEvaluation's multi_logloss: 1.17813\n",
      "[30]\tTrain's multi_logloss: 1.04808\tEvaluation's multi_logloss: 1.1678\n",
      "[31]\tTrain's multi_logloss: 1.03536\tEvaluation's multi_logloss: 1.15753\n",
      "[32]\tTrain's multi_logloss: 1.02327\tEvaluation's multi_logloss: 1.14803\n",
      "[33]\tTrain's multi_logloss: 1.01106\tEvaluation's multi_logloss: 1.13833\n",
      "[34]\tTrain's multi_logloss: 0.999807\tEvaluation's multi_logloss: 1.1294\n",
      "[35]\tTrain's multi_logloss: 0.988545\tEvaluation's multi_logloss: 1.12024\n",
      "[36]\tTrain's multi_logloss: 0.977613\tEvaluation's multi_logloss: 1.1117\n",
      "[37]\tTrain's multi_logloss: 0.967134\tEvaluation's multi_logloss: 1.1032\n",
      "[38]\tTrain's multi_logloss: 0.95707\tEvaluation's multi_logloss: 1.09522\n",
      "[39]\tTrain's multi_logloss: 0.947033\tEvaluation's multi_logloss: 1.08744\n",
      "[40]\tTrain's multi_logloss: 0.937215\tEvaluation's multi_logloss: 1.07944\n",
      "[41]\tTrain's multi_logloss: 0.927668\tEvaluation's multi_logloss: 1.07205\n",
      "[42]\tTrain's multi_logloss: 0.918192\tEvaluation's multi_logloss: 1.06502\n",
      "[43]\tTrain's multi_logloss: 0.909228\tEvaluation's multi_logloss: 1.05829\n",
      "[44]\tTrain's multi_logloss: 0.900302\tEvaluation's multi_logloss: 1.05164\n",
      "[45]\tTrain's multi_logloss: 0.891755\tEvaluation's multi_logloss: 1.04457\n",
      "[46]\tTrain's multi_logloss: 0.883389\tEvaluation's multi_logloss: 1.03802\n",
      "[47]\tTrain's multi_logloss: 0.875275\tEvaluation's multi_logloss: 1.03186\n",
      "[48]\tTrain's multi_logloss: 0.867355\tEvaluation's multi_logloss: 1.02566\n",
      "[49]\tTrain's multi_logloss: 0.859328\tEvaluation's multi_logloss: 1.01955\n",
      "[50]\tTrain's multi_logloss: 0.851845\tEvaluation's multi_logloss: 1.014\n",
      "[51]\tTrain's multi_logloss: 0.844479\tEvaluation's multi_logloss: 1.00848\n",
      "[52]\tTrain's multi_logloss: 0.83708\tEvaluation's multi_logloss: 1.00278\n",
      "[53]\tTrain's multi_logloss: 0.829948\tEvaluation's multi_logloss: 0.997262\n",
      "[54]\tTrain's multi_logloss: 0.823104\tEvaluation's multi_logloss: 0.992355\n",
      "[55]\tTrain's multi_logloss: 0.81631\tEvaluation's multi_logloss: 0.98733\n",
      "[56]\tTrain's multi_logloss: 0.809734\tEvaluation's multi_logloss: 0.982717\n",
      "[57]\tTrain's multi_logloss: 0.803235\tEvaluation's multi_logloss: 0.978005\n",
      "[58]\tTrain's multi_logloss: 0.796976\tEvaluation's multi_logloss: 0.973242\n",
      "[59]\tTrain's multi_logloss: 0.790825\tEvaluation's multi_logloss: 0.968687\n",
      "[60]\tTrain's multi_logloss: 0.785061\tEvaluation's multi_logloss: 0.964605\n",
      "[61]\tTrain's multi_logloss: 0.77938\tEvaluation's multi_logloss: 0.960673\n",
      "[62]\tTrain's multi_logloss: 0.773615\tEvaluation's multi_logloss: 0.956402\n",
      "[63]\tTrain's multi_logloss: 0.767983\tEvaluation's multi_logloss: 0.951957\n",
      "[64]\tTrain's multi_logloss: 0.762536\tEvaluation's multi_logloss: 0.948206\n",
      "[65]\tTrain's multi_logloss: 0.757162\tEvaluation's multi_logloss: 0.944395\n",
      "[66]\tTrain's multi_logloss: 0.751855\tEvaluation's multi_logloss: 0.940296\n",
      "[67]\tTrain's multi_logloss: 0.746736\tEvaluation's multi_logloss: 0.9364\n",
      "[68]\tTrain's multi_logloss: 0.741732\tEvaluation's multi_logloss: 0.932721\n",
      "[69]\tTrain's multi_logloss: 0.736692\tEvaluation's multi_logloss: 0.929286\n",
      "[70]\tTrain's multi_logloss: 0.731954\tEvaluation's multi_logloss: 0.925754\n",
      "[71]\tTrain's multi_logloss: 0.727203\tEvaluation's multi_logloss: 0.922195\n",
      "[72]\tTrain's multi_logloss: 0.722559\tEvaluation's multi_logloss: 0.918972\n",
      "[73]\tTrain's multi_logloss: 0.718124\tEvaluation's multi_logloss: 0.915597\n",
      "[74]\tTrain's multi_logloss: 0.713656\tEvaluation's multi_logloss: 0.912493\n",
      "[75]\tTrain's multi_logloss: 0.709337\tEvaluation's multi_logloss: 0.909172\n",
      "[76]\tTrain's multi_logloss: 0.70511\tEvaluation's multi_logloss: 0.906215\n",
      "[77]\tTrain's multi_logloss: 0.700836\tEvaluation's multi_logloss: 0.903323\n",
      "[78]\tTrain's multi_logloss: 0.696866\tEvaluation's multi_logloss: 0.900341\n",
      "[79]\tTrain's multi_logloss: 0.692945\tEvaluation's multi_logloss: 0.8976\n",
      "[80]\tTrain's multi_logloss: 0.688993\tEvaluation's multi_logloss: 0.894873\n",
      "[81]\tTrain's multi_logloss: 0.685118\tEvaluation's multi_logloss: 0.892058\n",
      "[82]\tTrain's multi_logloss: 0.68145\tEvaluation's multi_logloss: 0.889377\n",
      "[83]\tTrain's multi_logloss: 0.677648\tEvaluation's multi_logloss: 0.887026\n",
      "[84]\tTrain's multi_logloss: 0.673941\tEvaluation's multi_logloss: 0.884559\n",
      "[85]\tTrain's multi_logloss: 0.670644\tEvaluation's multi_logloss: 0.882175\n",
      "[86]\tTrain's multi_logloss: 0.667142\tEvaluation's multi_logloss: 0.879725\n",
      "[87]\tTrain's multi_logloss: 0.663698\tEvaluation's multi_logloss: 0.877377\n",
      "[88]\tTrain's multi_logloss: 0.660452\tEvaluation's multi_logloss: 0.875036\n",
      "[89]\tTrain's multi_logloss: 0.657222\tEvaluation's multi_logloss: 0.8728\n",
      "[90]\tTrain's multi_logloss: 0.65405\tEvaluation's multi_logloss: 0.870389\n",
      "[91]\tTrain's multi_logloss: 0.651032\tEvaluation's multi_logloss: 0.868091\n",
      "[92]\tTrain's multi_logloss: 0.64806\tEvaluation's multi_logloss: 0.865993\n",
      "[93]\tTrain's multi_logloss: 0.644948\tEvaluation's multi_logloss: 0.864205\n",
      "[94]\tTrain's multi_logloss: 0.641975\tEvaluation's multi_logloss: 0.86217\n",
      "[95]\tTrain's multi_logloss: 0.639162\tEvaluation's multi_logloss: 0.860391\n",
      "[96]\tTrain's multi_logloss: 0.636254\tEvaluation's multi_logloss: 0.858402\n",
      "[97]\tTrain's multi_logloss: 0.633508\tEvaluation's multi_logloss: 0.85632\n",
      "[98]\tTrain's multi_logloss: 0.63057\tEvaluation's multi_logloss: 0.854624\n",
      "[99]\tTrain's multi_logloss: 0.627835\tEvaluation's multi_logloss: 0.852793\n",
      "[100]\tTrain's multi_logloss: 0.625176\tEvaluation's multi_logloss: 0.851195\n",
      "Did not meet early stopping. Best iteration is:\n",
      "[100]\tTrain's multi_logloss: 0.625176\tEvaluation's multi_logloss: 0.851195\n"
     ]
    }
   ],
   "source": [
    "import lightgbm as lgb\n",
    "\n",
    "lgb_train = lgb.Dataset(trainX, trainY)\n",
    "lgb_test = lgb.Dataset(testX, testY)\n",
    "\n",
    "lgb_params = {\n",
    "    'boosting_type': 'goss',\n",
    "    'objective': 'multiclass',\n",
    "    'num_class': 13,\n",
    "    'metric': 'multi_logloss',\n",
    "    'max_depth' : 8,\n",
    "    'learning_rate': 0.03,\n",
    "    'early_stopping_round' : 10,\n",
    "    'top_rate' : 0.5,\n",
    "    'lambda_l1' : 6,\n",
    "    'lambda_l2' : 5,\n",
    "    'seed' : 32\n",
    "}\n",
    "\n",
    "gbm = lgb.train(params=lgb_params, train_set=lgb_train,\n",
    "                num_boost_round = 100, valid_sets = [lgb_test, lgb_train],\n",
    "               valid_names = ['Evaluation', 'Train'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.8061377245508982\n",
      "0.750373692077728\n",
      "0.9891000652138836\n",
      "0.9668724859106729\n"
     ]
    }
   ],
   "source": [
    "print(accuracy_score(trainY, np.argmax(gbm.predict(trainX), axis=1)))\n",
    "print(accuracy_score(testY, np.argmax(gbm.predict(testX), axis=1)))\n",
    "print(roc_auc_score(pd.get_dummies(trainY, prefix='label'), gbm.predict(trainX), average='micro'))\n",
    "print(roc_auc_score(pd.get_dummies(testY, prefix='label'), gbm.predict(testX), average='micro'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "import hyperopt as hp\n",
    "from hyperopt import Trials,fmin,STATUS_OK"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100%|██████████| 50/50 [16:08<00:00, 19.37s/it, best loss: 0.23007291892084403]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import cross_val_score\n",
    "\n",
    "rf_space = {\n",
    "    'n_estimators': hp.hp.choice('n_estimators', np.arange(50,500,1, dtype=int)),\n",
    "    'max_depth' : hp.hp.choice('max_depth', np.arange(1,50,1, dtype=int)),\n",
    "    'min_samples_split' : hp.hp.choice('min_samples_split', np.arange(1,100,1, dtype=int)),\n",
    "    'random_state' : 10\n",
    "    }\n",
    "\n",
    "def objective_m(params, n_folds=10):\n",
    "    model = RandomForestClassifier(**params)\n",
    "    score = cross_val_score(model, trainX, trainY, cv=n_folds, scoring='accuracy')\n",
    "    \n",
    "    loss = 1 - score.mean()\n",
    "    return loss\n",
    "\n",
    "bayes_trials = Trials()\n",
    "MAX_EVALS = 50 # this controls the runtime \n",
    "\n",
    "lgbm_best_m = fmin(fn=objective_m, space=rf_space, algo=hp.tpe.suggest, \n",
    "                   max_evals=MAX_EVALS, trials=bayes_trials)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'max_depth': 15, 'min_samples_split': 13, 'n_estimators': 38}"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lgbm_best_m"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',\n",
       "                       max_depth=7, max_features='auto', max_leaf_nodes=None,\n",
       "                       min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                       min_samples_leaf=1, min_samples_split=24,\n",
       "                       min_weight_fraction_leaf=0.0, n_estimators=70,\n",
       "                       n_jobs=None, oob_score=False, random_state=10, verbose=0,\n",
       "                       warm_start=False)"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf_b = RandomForestClassifier(n_estimators=70, max_depth=7, min_samples_split=24, random_state=10)\n",
    "rf_b.fit(trainX, trainY)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7440119760479041\n",
      "0.7249626307922272\n",
      "0.9825708449306477\n",
      "0.9677565367253478\n"
     ]
    }
   ],
   "source": [
    "print(accuracy_score(trainY, rf_b.predict(trainX)))\n",
    "print(accuracy_score(testY, rf_b.predict(testX)))\n",
    "print(roc_auc_score(pd.get_dummies(trainY, prefix='label'), rf_b.predict_proba(trainX), average='micro'))\n",
    "print(roc_auc_score(pd.get_dummies(testY, prefix='label'), rf_b.predict_proba(testX), average='micro'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import itertools\n",
    "from sklearn.metrics import confusion_matrix\n",
    "plt.style.use('seaborn')\n",
    "\n",
    "matrix1 = confusion_matrix(trainY, rf.predict(trainX))\n",
    "classes = list(label_dict.values())\n",
    "\n",
    "plt.imshow(matrix1,interpolation='nearest', cmap=plt.cm.Blues)\n",
    "plt.title('Confusion Matrix of Training Set')\n",
    "plt.colorbar()\n",
    "tick_marks = np.arange(len(classes))\n",
    "plt.xticks(tick_marks, classes, rotation=45)\n",
    "plt.yticks(tick_marks, classes)\n",
    "thresh = matrix1.max()/2.\n",
    "for i, j in itertools.product(range(matrix1.shape[0]), range(matrix1.shape[1])):\n",
    "    plt.text(j, i, format(matrix1[i, j], 'd'), horizontalalignment=\"center\",\n",
    "        color=\"white\" if matrix1[i, j] > thresh else \"black\")\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.ylabel('True Label')\n",
    "plt.xlabel('Predicted Label')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import roc_curve, roc_auc_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.style.use('seaborn')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "color_map = ['b', 'g', 'r', 'c', 'm', 'y', 'k', 'purple', 'orange', 'brown', 'pink', 'gray', 'olive']\n",
    "\n",
    "for i in range(len(label_dict)):\n",
    "    plt.plot(roc_curve((trainY==i), rf_b.predict_proba(trainX)[:,i])[0], roc_curve((trainY==i), \n",
    "            rf_b.predict_proba(trainX)[:,i])[1], color=color_map[i], \n",
    "            label=label_dict[i]+': ROC Curve (area = %0.5f)' % roc_auc_score((trainY==i), \n",
    "            rf_b.predict_proba(trainX)[:,i]))\n",
    "\n",
    "plt.plot([0, 2], [0, 2], color='black', linestyle='--')\n",
    "plt.xlim([0.0, 1.0])\n",
    "plt.ylim([0.0, 1.0])\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('RandomForest ROC-AUC on Training Set')\n",
    "plt.legend()\n",
    "#plt.show()\n",
    "plt.savefig('roc_train', dpi=500)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x396 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for i in range(len(label_dict)):\n",
    "    plt.plot(roc_curve((testY==i), rf_b.predict_proba(testX)[:,i])[0], roc_curve((testY==i), \n",
    "            rf_b.predict_proba(testX)[:,i])[1], color=color_map[i], \n",
    "            label=label_dict[i]+': ROC Curve (area = %0.5f)' % roc_auc_score((testY==i), \n",
    "            rf_b.predict_proba(testX)[:,i]))\n",
    "\n",
    "plt.plot([0, 2], [0, 2], color='black', linestyle='--')\n",
    "plt.xlim([0.0, 1.0])\n",
    "plt.ylim([0.0, 1.0])\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('RandomForest ROC-AUC on Test Set')\n",
    "plt.legend()\n",
    "#plt.show()\n",
    "plt.savefig('roc_test', dpi=500)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "import shap\n",
    "\n",
    "rf_shap_explainer = shap.TreeExplainer(rf_b)\n",
    "rf_shap_vals_train = rf_shap_explainer.shap_values(trainX)\n",
    "rf_shap_vals_test = rf_shap_explainer.shap_values(testX)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x684 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "shap.summary_plot(rf_shap_vals_train, trainX, show=False)\n",
    "plt.savefig('shap_train', dpi=500)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x684 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "shap.summary_plot(rf_shap_vals_test, testX, show=False)\n",
    "plt.savefig('shap_test', dpi=500)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_text = \"Zn(NO3)2·6H2O (0.297g 1 mmol) were dissolved in 70 mL of deionized water and sealed in the steel autoclave and heated into 120 °C. After the reaction, the autoclave was cooled to room temperature.\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "['Zn(NO3)2·6H2O', '0.297g', '1', 'mmol', 'were', 'dissolved', 'in', '70', 'mL', 'of', 'deionized', 'water', 'and', 'sealed', 'in', 'the', 'steel', 'autoclave', 'and', 'heated', 'into', '120', '°C', 'After', 'the', 'reaction', 'the', 'autoclave', 'was', 'cooled', 'to', 'room', 'temperature']"
      ],
      "text/plain": [
       "['Zn(NO3)2·6H2O',\n",
       " '0.297g',\n",
       " '1',\n",
       " 'mmol',\n",
       " 'were',\n",
       " 'dissolved',\n",
       " 'in',\n",
       " '70',\n",
       " 'mL',\n",
       " 'of',\n",
       " 'deionized',\n",
       " 'water',\n",
       " 'and',\n",
       " 'sealed',\n",
       " 'in',\n",
       " 'the',\n",
       " 'steel',\n",
       " 'autoclave',\n",
       " 'and',\n",
       " 'heated',\n",
       " 'into',\n",
       " '120',\n",
       " '°C',\n",
       " 'After',\n",
       " 'the',\n",
       " 'reaction',\n",
       " 'the',\n",
       " 'autoclave',\n",
       " 'was',\n",
       " 'cooled',\n",
       " 'to',\n",
       " 'room',\n",
       " 'temperature']"
      ]
     },
     "execution_count": 114,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_words = []\n",
    "tokens = test_text.split(' ')\n",
    "tokens = [word.strip(\"\\(\\).\\),\\[\\].∼;\") for word in tokens]\n",
    "test_words.extend(tokens)\n",
    "test_words"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "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>POS</th>\n",
       "      <th>CEM</th>\n",
       "      <th>isalpha</th>\n",
       "      <th>isdigit</th>\n",
       "      <th>isstop</th>\n",
       "      <th>charlen</th>\n",
       "      <th>0wordvec</th>\n",
       "      <th>1wordvec</th>\n",
       "      <th>2wordvec</th>\n",
       "      <th>3wordvec</th>\n",
       "      <th>4wordvec</th>\n",
       "      <th>5wordvec</th>\n",
       "      <th>6wordvec</th>\n",
       "      <th>7wordvec</th>\n",
       "      <th>8wordvec</th>\n",
       "      <th>9wordvec</th>\n",
       "      <th>10wordvec</th>\n",
       "      <th>11wordvec</th>\n",
       "      <th>12wordvec</th>\n",
       "      <th>13wordvec</th>\n",
       "      <th>14wordvec</th>\n",
       "      <th>15wordvec</th>\n",
       "      <th>16wordvec</th>\n",
       "      <th>17wordvec</th>\n",
       "      <th>18wordvec</th>\n",
       "      <th>19wordvec</th>\n",
       "      <th>20wordvec</th>\n",
       "      <th>21wordvec</th>\n",
       "      <th>22wordvec</th>\n",
       "      <th>23wordvec</th>\n",
       "      <th>24wordvec</th>\n",
       "      <th>25wordvec</th>\n",
       "      <th>26wordvec</th>\n",
       "      <th>27wordvec</th>\n",
       "      <th>28wordvec</th>\n",
       "      <th>29wordvec</th>\n",
       "      <th>30wordvec</th>\n",
       "      <th>31wordvec</th>\n",
       "      <th>32wordvec</th>\n",
       "      <th>33wordvec</th>\n",
       "      <th>34wordvec</th>\n",
       "      <th>35wordvec</th>\n",
       "      <th>36wordvec</th>\n",
       "      <th>37wordvec</th>\n",
       "      <th>38wordvec</th>\n",
       "      <th>39wordvec</th>\n",
       "      <th>40wordvec</th>\n",
       "      <th>41wordvec</th>\n",
       "      <th>42wordvec</th>\n",
       "      <th>43wordvec</th>\n",
       "      <th>44wordvec</th>\n",
       "      <th>45wordvec</th>\n",
       "      <th>46wordvec</th>\n",
       "      <th>47wordvec</th>\n",
       "      <th>48wordvec</th>\n",
       "      <th>49wordvec</th>\n",
       "      <th>50wordvec</th>\n",
       "      <th>51wordvec</th>\n",
       "      <th>52wordvec</th>\n",
       "      <th>53wordvec</th>\n",
       "      <th>54wordvec</th>\n",
       "      <th>55wordvec</th>\n",
       "      <th>56wordvec</th>\n",
       "      <th>57wordvec</th>\n",
       "      <th>58wordvec</th>\n",
       "      <th>59wordvec</th>\n",
       "      <th>60wordvec</th>\n",
       "      <th>61wordvec</th>\n",
       "      <th>62wordvec</th>\n",
       "      <th>63wordvec</th>\n",
       "      <th>64wordvec</th>\n",
       "      <th>65wordvec</th>\n",
       "      <th>66wordvec</th>\n",
       "      <th>67wordvec</th>\n",
       "      <th>68wordvec</th>\n",
       "      <th>69wordvec</th>\n",
       "      <th>70wordvec</th>\n",
       "      <th>71wordvec</th>\n",
       "      <th>72wordvec</th>\n",
       "      <th>73wordvec</th>\n",
       "      <th>74wordvec</th>\n",
       "      <th>75wordvec</th>\n",
       "      <th>76wordvec</th>\n",
       "      <th>77wordvec</th>\n",
       "      <th>78wordvec</th>\n",
       "      <th>79wordvec</th>\n",
       "      <th>80wordvec</th>\n",
       "      <th>81wordvec</th>\n",
       "      <th>82wordvec</th>\n",
       "      <th>83wordvec</th>\n",
       "      <th>84wordvec</th>\n",
       "      <th>85wordvec</th>\n",
       "      <th>86wordvec</th>\n",
       "      <th>87wordvec</th>\n",
       "      <th>88wordvec</th>\n",
       "      <th>89wordvec</th>\n",
       "      <th>90wordvec</th>\n",
       "      <th>91wordvec</th>\n",
       "      <th>92wordvec</th>\n",
       "      <th>93wordvec</th>\n",
       "      <th>94wordvec</th>\n",
       "      <th>95wordvec</th>\n",
       "      <th>96wordvec</th>\n",
       "      <th>97wordvec</th>\n",
       "      <th>98wordvec</th>\n",
       "      <th>99wordvec</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>13</td>\n",
       "      <td>2.971773</td>\n",
       "      <td>0.079782</td>\n",
       "      <td>-1.442658</td>\n",
       "      <td>0.906095</td>\n",
       "      <td>-0.578173</td>\n",
       "      <td>-3.456996</td>\n",
       "      <td>1.818855</td>\n",
       "      <td>-0.249961</td>\n",
       "      <td>-1.747058</td>\n",
       "      <td>2.538630</td>\n",
       "      <td>-1.408008</td>\n",
       "      <td>1.296400</td>\n",
       "      <td>-1.895300</td>\n",
       "      <td>1.388852</td>\n",
       "      <td>1.348856</td>\n",
       "      <td>-0.605443</td>\n",
       "      <td>-1.427501</td>\n",
       "      <td>-1.374920</td>\n",
       "      <td>0.979204</td>\n",
       "      <td>2.190984</td>\n",
       "      <td>1.321189</td>\n",
       "      <td>-2.331494</td>\n",
       "      <td>-0.300023</td>\n",
       "      <td>0.081007</td>\n",
       "      <td>1.284108</td>\n",
       "      <td>0.191856</td>\n",
       "      <td>-0.324322</td>\n",
       "      <td>-0.673941</td>\n",
       "      <td>-0.207134</td>\n",
       "      <td>1.835657</td>\n",
       "      <td>0.162647</td>\n",
       "      <td>1.817251</td>\n",
       "      <td>-1.771946</td>\n",
       "      <td>-0.801250</td>\n",
       "      <td>0.477380</td>\n",
       "      <td>0.664459</td>\n",
       "      <td>-0.160458</td>\n",
       "      <td>0.800591</td>\n",
       "      <td>0.601482</td>\n",
       "      <td>1.418731</td>\n",
       "      <td>-2.534163</td>\n",
       "      <td>0.691521</td>\n",
       "      <td>-1.534083</td>\n",
       "      <td>-0.779196</td>\n",
       "      <td>-1.024923</td>\n",
       "      <td>0.729365</td>\n",
       "      <td>2.298255</td>\n",
       "      <td>0.036849</td>\n",
       "      <td>-2.065656</td>\n",
       "      <td>-0.846711</td>\n",
       "      <td>0.339779</td>\n",
       "      <td>-0.837030</td>\n",
       "      <td>-3.363577</td>\n",
       "      <td>-1.336825</td>\n",
       "      <td>-1.028179</td>\n",
       "      <td>1.231302</td>\n",
       "      <td>-1.776891</td>\n",
       "      <td>-0.167647</td>\n",
       "      <td>0.312895</td>\n",
       "      <td>-0.773332</td>\n",
       "      <td>-0.349892</td>\n",
       "      <td>2.225868</td>\n",
       "      <td>2.763717</td>\n",
       "      <td>-1.292496</td>\n",
       "      <td>2.191343</td>\n",
       "      <td>-1.227325</td>\n",
       "      <td>0.622864</td>\n",
       "      <td>0.708385</td>\n",
       "      <td>-1.176783</td>\n",
       "      <td>-0.813089</td>\n",
       "      <td>-0.460365</td>\n",
       "      <td>-1.728312</td>\n",
       "      <td>-1.776993</td>\n",
       "      <td>-0.967790</td>\n",
       "      <td>-0.537336</td>\n",
       "      <td>2.094493</td>\n",
       "      <td>-4.028937</td>\n",
       "      <td>-0.010159</td>\n",
       "      <td>-0.342716</td>\n",
       "      <td>1.329443</td>\n",
       "      <td>0.747531</td>\n",
       "      <td>3.381051</td>\n",
       "      <td>-0.176453</td>\n",
       "      <td>-0.763359</td>\n",
       "      <td>3.472584</td>\n",
       "      <td>0.261220</td>\n",
       "      <td>-2.064681</td>\n",
       "      <td>2.734037</td>\n",
       "      <td>0.619349</td>\n",
       "      <td>-1.602731</td>\n",
       "      <td>-1.050604</td>\n",
       "      <td>1.175166</td>\n",
       "      <td>-1.393727</td>\n",
       "      <td>0.011051</td>\n",
       "      <td>0.272701</td>\n",
       "      <td>0.308361</td>\n",
       "      <td>1.882590</td>\n",
       "      <td>-0.065714</td>\n",
       "      <td>0.252366</td>\n",
       "      <td>-1.437738</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.396979</td>\n",
       "      <td>0.937799</td>\n",
       "      <td>0.271259</td>\n",
       "      <td>-0.050497</td>\n",
       "      <td>0.911645</td>\n",
       "      <td>-1.025809</td>\n",
       "      <td>0.996646</td>\n",
       "      <td>0.422837</td>\n",
       "      <td>0.237371</td>\n",
       "      <td>1.246727</td>\n",
       "      <td>-0.942901</td>\n",
       "      <td>0.790358</td>\n",
       "      <td>-0.795013</td>\n",
       "      <td>1.271904</td>\n",
       "      <td>-0.722416</td>\n",
       "      <td>-0.178237</td>\n",
       "      <td>-1.586349</td>\n",
       "      <td>0.865627</td>\n",
       "      <td>0.692665</td>\n",
       "      <td>-1.412019</td>\n",
       "      <td>1.823491</td>\n",
       "      <td>1.386488</td>\n",
       "      <td>0.380556</td>\n",
       "      <td>0.516361</td>\n",
       "      <td>1.130560</td>\n",
       "      <td>-0.788502</td>\n",
       "      <td>0.020407</td>\n",
       "      <td>-0.303782</td>\n",
       "      <td>1.502163</td>\n",
       "      <td>1.013205</td>\n",
       "      <td>-1.710870</td>\n",
       "      <td>-0.486070</td>\n",
       "      <td>-0.345496</td>\n",
       "      <td>0.194273</td>\n",
       "      <td>0.979149</td>\n",
       "      <td>0.840505</td>\n",
       "      <td>0.949236</td>\n",
       "      <td>-0.812165</td>\n",
       "      <td>0.700657</td>\n",
       "      <td>0.828382</td>\n",
       "      <td>-1.742127</td>\n",
       "      <td>-2.161091</td>\n",
       "      <td>-0.392913</td>\n",
       "      <td>-0.305685</td>\n",
       "      <td>-1.137358</td>\n",
       "      <td>1.179231</td>\n",
       "      <td>-0.200112</td>\n",
       "      <td>-1.120268</td>\n",
       "      <td>1.604718</td>\n",
       "      <td>-0.854580</td>\n",
       "      <td>1.062583</td>\n",
       "      <td>0.546246</td>\n",
       "      <td>-2.102802</td>\n",
       "      <td>-1.587746</td>\n",
       "      <td>0.813604</td>\n",
       "      <td>0.280103</td>\n",
       "      <td>1.650809</td>\n",
       "      <td>-0.384675</td>\n",
       "      <td>-1.189358</td>\n",
       "      <td>1.579419</td>\n",
       "      <td>0.382736</td>\n",
       "      <td>-0.201886</td>\n",
       "      <td>0.571785</td>\n",
       "      <td>-0.847450</td>\n",
       "      <td>1.160882</td>\n",
       "      <td>-1.258801</td>\n",
       "      <td>-0.581712</td>\n",
       "      <td>-1.093447</td>\n",
       "      <td>1.151655</td>\n",
       "      <td>0.351244</td>\n",
       "      <td>-0.579111</td>\n",
       "      <td>-0.972813</td>\n",
       "      <td>-0.347316</td>\n",
       "      <td>0.127437</td>\n",
       "      <td>-0.119295</td>\n",
       "      <td>0.943783</td>\n",
       "      <td>1.954754</td>\n",
       "      <td>-1.146129</td>\n",
       "      <td>0.707329</td>\n",
       "      <td>-0.335946</td>\n",
       "      <td>-0.342549</td>\n",
       "      <td>0.170171</td>\n",
       "      <td>-0.196283</td>\n",
       "      <td>-1.371183</td>\n",
       "      <td>0.848139</td>\n",
       "      <td>0.942557</td>\n",
       "      <td>-1.635065</td>\n",
       "      <td>-0.861431</td>\n",
       "      <td>-1.682000</td>\n",
       "      <td>0.122985</td>\n",
       "      <td>0.437747</td>\n",
       "      <td>-1.123901</td>\n",
       "      <td>-0.849618</td>\n",
       "      <td>-0.357624</td>\n",
       "      <td>-1.710191</td>\n",
       "      <td>0.523278</td>\n",
       "      <td>1.051634</td>\n",
       "      <td>0.275828</td>\n",
       "      <td>-0.832872</td>\n",
       "      <td>-0.193382</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>-1.867648</td>\n",
       "      <td>0.491826</td>\n",
       "      <td>0.293565</td>\n",
       "      <td>2.133936</td>\n",
       "      <td>-4.002116</td>\n",
       "      <td>-1.528622</td>\n",
       "      <td>2.824345</td>\n",
       "      <td>-0.274839</td>\n",
       "      <td>-1.665203</td>\n",
       "      <td>2.042281</td>\n",
       "      <td>-0.586950</td>\n",
       "      <td>-0.696676</td>\n",
       "      <td>1.594625</td>\n",
       "      <td>3.787341</td>\n",
       "      <td>-1.717083</td>\n",
       "      <td>1.893366</td>\n",
       "      <td>0.828534</td>\n",
       "      <td>-2.297643</td>\n",
       "      <td>-4.099036</td>\n",
       "      <td>-1.352220</td>\n",
       "      <td>1.548374</td>\n",
       "      <td>-2.478937</td>\n",
       "      <td>1.338818</td>\n",
       "      <td>1.579769</td>\n",
       "      <td>4.725752</td>\n",
       "      <td>-0.895735</td>\n",
       "      <td>-1.631403</td>\n",
       "      <td>-1.083338</td>\n",
       "      <td>-1.281868</td>\n",
       "      <td>2.343051</td>\n",
       "      <td>2.456086</td>\n",
       "      <td>1.972860</td>\n",
       "      <td>-1.290680</td>\n",
       "      <td>2.656415</td>\n",
       "      <td>-4.970997</td>\n",
       "      <td>-2.044782</td>\n",
       "      <td>-1.257914</td>\n",
       "      <td>1.558470</td>\n",
       "      <td>-0.506398</td>\n",
       "      <td>-1.503979</td>\n",
       "      <td>3.601033</td>\n",
       "      <td>3.645257</td>\n",
       "      <td>-1.544224</td>\n",
       "      <td>3.626883</td>\n",
       "      <td>-0.313149</td>\n",
       "      <td>-1.378304</td>\n",
       "      <td>1.731402</td>\n",
       "      <td>-0.605043</td>\n",
       "      <td>-6.303889</td>\n",
       "      <td>4.450224</td>\n",
       "      <td>-0.729304</td>\n",
       "      <td>-0.829006</td>\n",
       "      <td>-1.144873</td>\n",
       "      <td>0.077416</td>\n",
       "      <td>-0.263675</td>\n",
       "      <td>2.282919</td>\n",
       "      <td>-1.876464</td>\n",
       "      <td>-0.634014</td>\n",
       "      <td>1.204571</td>\n",
       "      <td>-2.829306</td>\n",
       "      <td>0.002334</td>\n",
       "      <td>2.430563</td>\n",
       "      <td>2.383279</td>\n",
       "      <td>0.230914</td>\n",
       "      <td>2.619345</td>\n",
       "      <td>-2.236077</td>\n",
       "      <td>-1.620061</td>\n",
       "      <td>4.455505</td>\n",
       "      <td>-2.311388</td>\n",
       "      <td>0.970543</td>\n",
       "      <td>1.190882</td>\n",
       "      <td>-1.561137</td>\n",
       "      <td>-2.142201</td>\n",
       "      <td>-1.894182</td>\n",
       "      <td>2.427807</td>\n",
       "      <td>-0.664178</td>\n",
       "      <td>-7.166413</td>\n",
       "      <td>1.353819</td>\n",
       "      <td>1.678267</td>\n",
       "      <td>-0.284935</td>\n",
       "      <td>0.056792</td>\n",
       "      <td>1.416663</td>\n",
       "      <td>-1.399460</td>\n",
       "      <td>0.605700</td>\n",
       "      <td>3.220578</td>\n",
       "      <td>-2.735284</td>\n",
       "      <td>-1.809777</td>\n",
       "      <td>0.946542</td>\n",
       "      <td>1.671572</td>\n",
       "      <td>-2.739539</td>\n",
       "      <td>-2.793553</td>\n",
       "      <td>-0.012796</td>\n",
       "      <td>-2.935273</td>\n",
       "      <td>0.835032</td>\n",
       "      <td>0.669231</td>\n",
       "      <td>1.841992</td>\n",
       "      <td>0.391523</td>\n",
       "      <td>0.873473</td>\n",
       "      <td>-2.468980</td>\n",
       "      <td>-2.104712</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>-0.194317</td>\n",
       "      <td>-1.353077</td>\n",
       "      <td>-4.256432</td>\n",
       "      <td>0.482153</td>\n",
       "      <td>1.231465</td>\n",
       "      <td>-3.485644</td>\n",
       "      <td>0.229919</td>\n",
       "      <td>-1.602212</td>\n",
       "      <td>0.487276</td>\n",
       "      <td>1.473904</td>\n",
       "      <td>-0.719897</td>\n",
       "      <td>1.512355</td>\n",
       "      <td>-0.633915</td>\n",
       "      <td>0.719186</td>\n",
       "      <td>-1.333272</td>\n",
       "      <td>-1.461292</td>\n",
       "      <td>-1.052577</td>\n",
       "      <td>2.104692</td>\n",
       "      <td>-0.140655</td>\n",
       "      <td>0.718868</td>\n",
       "      <td>-0.188677</td>\n",
       "      <td>-0.197250</td>\n",
       "      <td>1.252481</td>\n",
       "      <td>-1.925643</td>\n",
       "      <td>-0.031747</td>\n",
       "      <td>1.552012</td>\n",
       "      <td>-2.273719</td>\n",
       "      <td>-1.052019</td>\n",
       "      <td>0.710813</td>\n",
       "      <td>-2.876308</td>\n",
       "      <td>0.244386</td>\n",
       "      <td>-0.884932</td>\n",
       "      <td>-0.056248</td>\n",
       "      <td>-0.232628</td>\n",
       "      <td>-0.741219</td>\n",
       "      <td>-1.457794</td>\n",
       "      <td>-1.103274</td>\n",
       "      <td>-0.285785</td>\n",
       "      <td>-0.642462</td>\n",
       "      <td>-0.660129</td>\n",
       "      <td>1.321474</td>\n",
       "      <td>-2.190253</td>\n",
       "      <td>-0.917183</td>\n",
       "      <td>-0.453435</td>\n",
       "      <td>3.019042</td>\n",
       "      <td>-0.222213</td>\n",
       "      <td>2.235376</td>\n",
       "      <td>1.624835</td>\n",
       "      <td>0.257305</td>\n",
       "      <td>-4.249274</td>\n",
       "      <td>0.031026</td>\n",
       "      <td>-1.135420</td>\n",
       "      <td>1.718721</td>\n",
       "      <td>0.286074</td>\n",
       "      <td>0.679326</td>\n",
       "      <td>-3.588328</td>\n",
       "      <td>3.106425</td>\n",
       "      <td>2.748616</td>\n",
       "      <td>-1.744928</td>\n",
       "      <td>-0.172806</td>\n",
       "      <td>-1.048486</td>\n",
       "      <td>1.934354</td>\n",
       "      <td>0.446806</td>\n",
       "      <td>-0.694699</td>\n",
       "      <td>-1.639181</td>\n",
       "      <td>2.374921</td>\n",
       "      <td>0.085974</td>\n",
       "      <td>-1.143476</td>\n",
       "      <td>0.035740</td>\n",
       "      <td>-0.399286</td>\n",
       "      <td>1.151743</td>\n",
       "      <td>3.491725</td>\n",
       "      <td>0.075098</td>\n",
       "      <td>-0.937093</td>\n",
       "      <td>-0.115973</td>\n",
       "      <td>-0.397789</td>\n",
       "      <td>-0.510826</td>\n",
       "      <td>-0.802052</td>\n",
       "      <td>1.119894</td>\n",
       "      <td>2.689321</td>\n",
       "      <td>-1.574555</td>\n",
       "      <td>-0.918702</td>\n",
       "      <td>-1.288551</td>\n",
       "      <td>-0.474045</td>\n",
       "      <td>0.131825</td>\n",
       "      <td>-0.340348</td>\n",
       "      <td>-0.157096</td>\n",
       "      <td>-0.554358</td>\n",
       "      <td>-2.003514</td>\n",
       "      <td>0.429778</td>\n",
       "      <td>-0.191816</td>\n",
       "      <td>1.013996</td>\n",
       "      <td>-0.729612</td>\n",
       "      <td>3.914708</td>\n",
       "      <td>-0.099635</td>\n",
       "      <td>-0.657804</td>\n",
       "      <td>-1.070428</td>\n",
       "      <td>0.152561</td>\n",
       "      <td>-0.922177</td>\n",
       "      <td>-3.295008</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   POS  CEM  isalpha  isdigit  isstop  charlen  0wordvec  1wordvec  2wordvec  \\\n",
       "0    5    0        0        0       0       13  2.971773  0.079782 -1.442658   \n",
       "1    5    0        0        0       0        6  0.000000  0.000000  0.000000   \n",
       "2    1    0        0        1       0        1  0.396979  0.937799  0.271259   \n",
       "3    5    0        1        0       0        4 -1.867648  0.491826  0.293565   \n",
       "4    6    0        1        0       1        4 -0.194317 -1.353077 -4.256432   \n",
       "\n",
       "   3wordvec  4wordvec  5wordvec  6wordvec  7wordvec  8wordvec  9wordvec  \\\n",
       "0  0.906095 -0.578173 -3.456996  1.818855 -0.249961 -1.747058  2.538630   \n",
       "1  0.000000  0.000000  0.000000  0.000000  0.000000  0.000000  0.000000   \n",
       "2 -0.050497  0.911645 -1.025809  0.996646  0.422837  0.237371  1.246727   \n",
       "3  2.133936 -4.002116 -1.528622  2.824345 -0.274839 -1.665203  2.042281   \n",
       "4  0.482153  1.231465 -3.485644  0.229919 -1.602212  0.487276  1.473904   \n",
       "\n",
       "   10wordvec  11wordvec  12wordvec  13wordvec  14wordvec  15wordvec  \\\n",
       "0  -1.408008   1.296400  -1.895300   1.388852   1.348856  -0.605443   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2  -0.942901   0.790358  -0.795013   1.271904  -0.722416  -0.178237   \n",
       "3  -0.586950  -0.696676   1.594625   3.787341  -1.717083   1.893366   \n",
       "4  -0.719897   1.512355  -0.633915   0.719186  -1.333272  -1.461292   \n",
       "\n",
       "   16wordvec  17wordvec  18wordvec  19wordvec  20wordvec  21wordvec  \\\n",
       "0  -1.427501  -1.374920   0.979204   2.190984   1.321189  -2.331494   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2  -1.586349   0.865627   0.692665  -1.412019   1.823491   1.386488   \n",
       "3   0.828534  -2.297643  -4.099036  -1.352220   1.548374  -2.478937   \n",
       "4  -1.052577   2.104692  -0.140655   0.718868  -0.188677  -0.197250   \n",
       "\n",
       "   22wordvec  23wordvec  24wordvec  25wordvec  26wordvec  27wordvec  \\\n",
       "0  -0.300023   0.081007   1.284108   0.191856  -0.324322  -0.673941   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2   0.380556   0.516361   1.130560  -0.788502   0.020407  -0.303782   \n",
       "3   1.338818   1.579769   4.725752  -0.895735  -1.631403  -1.083338   \n",
       "4   1.252481  -1.925643  -0.031747   1.552012  -2.273719  -1.052019   \n",
       "\n",
       "   28wordvec  29wordvec  30wordvec  31wordvec  32wordvec  33wordvec  \\\n",
       "0  -0.207134   1.835657   0.162647   1.817251  -1.771946  -0.801250   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2   1.502163   1.013205  -1.710870  -0.486070  -0.345496   0.194273   \n",
       "3  -1.281868   2.343051   2.456086   1.972860  -1.290680   2.656415   \n",
       "4   0.710813  -2.876308   0.244386  -0.884932  -0.056248  -0.232628   \n",
       "\n",
       "   34wordvec  35wordvec  36wordvec  37wordvec  38wordvec  39wordvec  \\\n",
       "0   0.477380   0.664459  -0.160458   0.800591   0.601482   1.418731   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2   0.979149   0.840505   0.949236  -0.812165   0.700657   0.828382   \n",
       "3  -4.970997  -2.044782  -1.257914   1.558470  -0.506398  -1.503979   \n",
       "4  -0.741219  -1.457794  -1.103274  -0.285785  -0.642462  -0.660129   \n",
       "\n",
       "   40wordvec  41wordvec  42wordvec  43wordvec  44wordvec  45wordvec  \\\n",
       "0  -2.534163   0.691521  -1.534083  -0.779196  -1.024923   0.729365   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2  -1.742127  -2.161091  -0.392913  -0.305685  -1.137358   1.179231   \n",
       "3   3.601033   3.645257  -1.544224   3.626883  -0.313149  -1.378304   \n",
       "4   1.321474  -2.190253  -0.917183  -0.453435   3.019042  -0.222213   \n",
       "\n",
       "   46wordvec  47wordvec  48wordvec  49wordvec  50wordvec  51wordvec  \\\n",
       "0   2.298255   0.036849  -2.065656  -0.846711   0.339779  -0.837030   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2  -0.200112  -1.120268   1.604718  -0.854580   1.062583   0.546246   \n",
       "3   1.731402  -0.605043  -6.303889   4.450224  -0.729304  -0.829006   \n",
       "4   2.235376   1.624835   0.257305  -4.249274   0.031026  -1.135420   \n",
       "\n",
       "   52wordvec  53wordvec  54wordvec  55wordvec  56wordvec  57wordvec  \\\n",
       "0  -3.363577  -1.336825  -1.028179   1.231302  -1.776891  -0.167647   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2  -2.102802  -1.587746   0.813604   0.280103   1.650809  -0.384675   \n",
       "3  -1.144873   0.077416  -0.263675   2.282919  -1.876464  -0.634014   \n",
       "4   1.718721   0.286074   0.679326  -3.588328   3.106425   2.748616   \n",
       "\n",
       "   58wordvec  59wordvec  60wordvec  61wordvec  62wordvec  63wordvec  \\\n",
       "0   0.312895  -0.773332  -0.349892   2.225868   2.763717  -1.292496   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2  -1.189358   1.579419   0.382736  -0.201886   0.571785  -0.847450   \n",
       "3   1.204571  -2.829306   0.002334   2.430563   2.383279   0.230914   \n",
       "4  -1.744928  -0.172806  -1.048486   1.934354   0.446806  -0.694699   \n",
       "\n",
       "   64wordvec  65wordvec  66wordvec  67wordvec  68wordvec  69wordvec  \\\n",
       "0   2.191343  -1.227325   0.622864   0.708385  -1.176783  -0.813089   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2   1.160882  -1.258801  -0.581712  -1.093447   1.151655   0.351244   \n",
       "3   2.619345  -2.236077  -1.620061   4.455505  -2.311388   0.970543   \n",
       "4  -1.639181   2.374921   0.085974  -1.143476   0.035740  -0.399286   \n",
       "\n",
       "   70wordvec  71wordvec  72wordvec  73wordvec  74wordvec  75wordvec  \\\n",
       "0  -0.460365  -1.728312  -1.776993  -0.967790  -0.537336   2.094493   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2  -0.579111  -0.972813  -0.347316   0.127437  -0.119295   0.943783   \n",
       "3   1.190882  -1.561137  -2.142201  -1.894182   2.427807  -0.664178   \n",
       "4   1.151743   3.491725   0.075098  -0.937093  -0.115973  -0.397789   \n",
       "\n",
       "   76wordvec  77wordvec  78wordvec  79wordvec  80wordvec  81wordvec  \\\n",
       "0  -4.028937  -0.010159  -0.342716   1.329443   0.747531   3.381051   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2   1.954754  -1.146129   0.707329  -0.335946  -0.342549   0.170171   \n",
       "3  -7.166413   1.353819   1.678267  -0.284935   0.056792   1.416663   \n",
       "4  -0.510826  -0.802052   1.119894   2.689321  -1.574555  -0.918702   \n",
       "\n",
       "   82wordvec  83wordvec  84wordvec  85wordvec  86wordvec  87wordvec  \\\n",
       "0  -0.176453  -0.763359   3.472584   0.261220  -2.064681   2.734037   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2  -0.196283  -1.371183   0.848139   0.942557  -1.635065  -0.861431   \n",
       "3  -1.399460   0.605700   3.220578  -2.735284  -1.809777   0.946542   \n",
       "4  -1.288551  -0.474045   0.131825  -0.340348  -0.157096  -0.554358   \n",
       "\n",
       "   88wordvec  89wordvec  90wordvec  91wordvec  92wordvec  93wordvec  \\\n",
       "0   0.619349  -1.602731  -1.050604   1.175166  -1.393727   0.011051   \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000   \n",
       "2  -1.682000   0.122985   0.437747  -1.123901  -0.849618  -0.357624   \n",
       "3   1.671572  -2.739539  -2.793553  -0.012796  -2.935273   0.835032   \n",
       "4  -2.003514   0.429778  -0.191816   1.013996  -0.729612   3.914708   \n",
       "\n",
       "   94wordvec  95wordvec  96wordvec  97wordvec  98wordvec  99wordvec  \n",
       "0   0.272701   0.308361   1.882590  -0.065714   0.252366  -1.437738  \n",
       "1   0.000000   0.000000   0.000000   0.000000   0.000000   0.000000  \n",
       "2  -1.710191   0.523278   1.051634   0.275828  -0.832872  -0.193382  \n",
       "3   0.669231   1.841992   0.391523   0.873473  -2.468980  -2.104712  \n",
       "4  -0.099635  -0.657804  -1.070428   0.152561  -0.922177  -3.295008  "
      ]
     },
     "execution_count": 115,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_df = pd.DataFrame(test_words, columns=['word'])\n",
    "\n",
    "test_df['POS'] = [tag[1] for tag in cpt.tag(test_words)]\n",
    "test_df['CEM'] = [tag[1] for tag in cct.tag(test_words)]\n",
    "\n",
    "word_vec_list = np.zeros([len(test_words), 100])\n",
    "for i, word in enumerate(test_words):\n",
    "    try:\n",
    "        wordvec = model.wv.get_vector(word)\n",
    "        word_vec_list[i] = wordvec\n",
    "    except:\n",
    "        continue\n",
    "test_df['isalpha'] = [word.isalpha() for word in test_words]\n",
    "test_df['isdigit'] = [word.isdigit() for word in test_words]\n",
    "test_df['isstop'] = [word.lower() in stop for word in test_words]\n",
    "test_df['charlen'] = [len(word) for word in test_words]\n",
    "\n",
    "test_df = pd.concat([test_df.reset_index(drop=True), pd.DataFrame(word_vec_list).add_suffix('wordvec')], axis=1)\n",
    "for col in test_df.columns[1:6]:\n",
    "    test_df[col] = test_df[col].astype('category').cat.codes\n",
    "test_df['CEM'] = test_df['CEM'] + 1\n",
    "test_df.drop('word', axis=1, inplace=True)\n",
    "test_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {},
   "outputs": [],
   "source": [
    "res_df = pd.DataFrame({'word': test_words, 'label':[label_dict[pred] for pred in rf.predict(test_df)]})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "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>10</th>\n",
       "      <th>11</th>\n",
       "      <th>12</th>\n",
       "      <th>13</th>\n",
       "      <th>14</th>\n",
       "      <th>15</th>\n",
       "      <th>16</th>\n",
       "      <th>17</th>\n",
       "      <th>18</th>\n",
       "      <th>19</th>\n",
       "      <th>20</th>\n",
       "      <th>21</th>\n",
       "      <th>22</th>\n",
       "      <th>23</th>\n",
       "      <th>24</th>\n",
       "      <th>25</th>\n",
       "      <th>26</th>\n",
       "      <th>27</th>\n",
       "      <th>28</th>\n",
       "      <th>29</th>\n",
       "      <th>30</th>\n",
       "      <th>31</th>\n",
       "      <th>32</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>word</th>\n",
       "      <td>Zn(NO3)2·6H2O</td>\n",
       "      <td>0.297g</td>\n",
       "      <td>1</td>\n",
       "      <td>mmol</td>\n",
       "      <td>were</td>\n",
       "      <td>dissolved</td>\n",
       "      <td>in</td>\n",
       "      <td>70</td>\n",
       "      <td>mL</td>\n",
       "      <td>of</td>\n",
       "      <td>deionized</td>\n",
       "      <td>water</td>\n",
       "      <td>and</td>\n",
       "      <td>sealed</td>\n",
       "      <td>in</td>\n",
       "      <td>the</td>\n",
       "      <td>steel</td>\n",
       "      <td>autoclave</td>\n",
       "      <td>and</td>\n",
       "      <td>heated</td>\n",
       "      <td>into</td>\n",
       "      <td>120</td>\n",
       "      <td>°C</td>\n",
       "      <td>After</td>\n",
       "      <td>the</td>\n",
       "      <td>reaction</td>\n",
       "      <td>the</td>\n",
       "      <td>autoclave</td>\n",
       "      <td>was</td>\n",
       "      <td>cooled</td>\n",
       "      <td>to</td>\n",
       "      <td>room</td>\n",
       "      <td>temperature</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>label</th>\n",
       "      <td>reactant</td>\n",
       "      <td>null</td>\n",
       "      <td>number</td>\n",
       "      <td>amount unit</td>\n",
       "      <td>null</td>\n",
       "      <td>operation</td>\n",
       "      <td>null</td>\n",
       "      <td>number</td>\n",
       "      <td>amount unit</td>\n",
       "      <td>null</td>\n",
       "      <td>null</td>\n",
       "      <td>condition</td>\n",
       "      <td>null</td>\n",
       "      <td>null</td>\n",
       "      <td>null</td>\n",
       "      <td>null</td>\n",
       "      <td>null</td>\n",
       "      <td>apparatus</td>\n",
       "      <td>null</td>\n",
       "      <td>operation</td>\n",
       "      <td>null</td>\n",
       "      <td>number</td>\n",
       "      <td>condition unit</td>\n",
       "      <td>null</td>\n",
       "      <td>null</td>\n",
       "      <td>null</td>\n",
       "      <td>null</td>\n",
       "      <td>apparatus</td>\n",
       "      <td>null</td>\n",
       "      <td>operation</td>\n",
       "      <td>null</td>\n",
       "      <td>condition</td>\n",
       "      <td>condition</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   0       1       2            3     4          5     6  \\\n",
       "word   Zn(NO3)2·6H2O  0.297g       1         mmol  were  dissolved    in   \n",
       "label       reactant    null  number  amount unit  null  operation  null   \n",
       "\n",
       "            7            8     9         10         11    12      13    14  \\\n",
       "word       70           mL    of  deionized      water   and  sealed    in   \n",
       "label  number  amount unit  null       null  condition  null    null  null   \n",
       "\n",
       "         15     16         17    18         19    20      21              22  \\\n",
       "word    the  steel  autoclave   and     heated  into     120              °C   \n",
       "label  null   null  apparatus  null  operation  null  number  condition unit   \n",
       "\n",
       "          23    24        25    26         27    28         29    30  \\\n",
       "word   After   the  reaction   the  autoclave   was     cooled    to   \n",
       "label   null  null      null  null  apparatus  null  operation  null   \n",
       "\n",
       "              31           32  \n",
       "word        room  temperature  \n",
       "label  condition    condition  "
      ]
     },
     "execution_count": 124,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "˚"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}