feature_engineering_book.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "「機械学習のための特徴量エンジニアリング」の9章の動作確認\n",
    "\n",
    "メモ\n",
    "\n",
    "- データは https://www.openacademic.ai/oag/\n",
    "- ただしv1の方を使わないと動かない\n",
    "- メモリ32GByteでも足りない"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{\"title\": \"System and Method for Maskless Direct Write Lithography\", \"lang\": \"en\", \"year\": 2015, \"references\": [\"354c172f-d877-4e60-a7eb-c1b1cf03ce4d\", \"76cf1064-b2b2-4245-940b-4e25dab9d41d\"], \"abstract\": \"A system and method for maskless direct write lithography are disclosed. The method includes receiving a plurality of pixels that represent an integrated circuit (IC) layout; identifying a first subset of the pixels that are suitable for a first compression method; and identifying a second subset of the pixels that are suitable for a second compression method. The method further includes compressing the first and second subset using the first and second compression method respectively, resulting in compressed data. The method further includes delivering the compressed data to a maskless direct writer for manufacturing a substrate. In embodiments, the first compression method uses a run-length encoding and the second compression method uses a dictionary-based encoding. Due to the hybrid compression method, the compressed data can be decompressed with a data rate expansion ratio sufficient for high-volume IC manufacturing.\", \"url\": [\"http://www.freepatentsonline.com/y2016/0211117.html\", \"http://www.google.com/patents/US20160211117\", \"https://www.google.de/patents/US20160211117\"], \"id\": \"0000002e-c2f2-4e25-9341-60d39130ac7a\", \"fos\": [\"Electronic engineering\", \"Computer hardware\", \"Engineering\", \"Engineering drawing\"]}\r\n"
     ]
    }
   ],
   "source": [
    "!head ./data/mag_papers_0.txt -n 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000000 ./data/mag_papers_0.txt\r\n"
     ]
    }
   ],
   "source": [
    "!wc -l ./data/mag_papers_0.txt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "!head ./data/mag_papers_0.txt -n 20000 > ./data/mag_papers_0_head.txt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(20000, 19)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "orig_df = pd.read_json('./data/mag_papers_0_head.txt', lines=True)\n",
    "orig_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>abstract</th>\n",
       "      <th>authors</th>\n",
       "      <th>doc_type</th>\n",
       "      <th>doi</th>\n",
       "      <th>fos</th>\n",
       "      <th>id</th>\n",
       "      <th>issue</th>\n",
       "      <th>keywords</th>\n",
       "      <th>lang</th>\n",
       "      <th>n_citation</th>\n",
       "      <th>page_end</th>\n",
       "      <th>page_start</th>\n",
       "      <th>publisher</th>\n",
       "      <th>references</th>\n",
       "      <th>title</th>\n",
       "      <th>url</th>\n",
       "      <th>venue</th>\n",
       "      <th>volume</th>\n",
       "      <th>year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A system and method for maskless direct write ...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>[Electronic engineering, Computer hardware, En...</td>\n",
       "      <td>0000002e-c2f2-4e25-9341-60d39130ac7a</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>en</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>[354c172f-d877-4e60-a7eb-c1b1cf03ce4d, 76cf106...</td>\n",
       "      <td>System and Method for Maskless Direct Write Li...</td>\n",
       "      <td>[http://www.freepatentsonline.com/y2016/021111...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>NaN</td>\n",
       "      <td>[{'name': 'Ahmed M. Alluwaimi'}]</td>\n",
       "      <td>NaN</td>\n",
       "      <td>10.4172/2157-7560.C1.030</td>\n",
       "      <td>[Biology, Virology, Immunology, Microbiology]</td>\n",
       "      <td>0000003d-5ce0-4eed-a56f-bbd3f8d8242b</td>\n",
       "      <td>NaN</td>\n",
       "      <td>[paratuberculosis, of, subspecies, proceedings...</td>\n",
       "      <td>en</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>OMICS International</td>\n",
       "      <td>NaN</td>\n",
       "      <td>The dilemma of the Mycobacterium avium subspec...</td>\n",
       "      <td>[http://www.omicsonline.org/proceedings/the-di...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NaN</td>\n",
       "      <td>[{'name': 'Jovana P. Lekovich', 'org': 'Weill ...</td>\n",
       "      <td>Journal</td>\n",
       "      <td>10.1016/j.fertnstert.2015.07.492</td>\n",
       "      <td>NaN</td>\n",
       "      <td>00000061-0bab-468b-9481-de0c899783f3</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>en</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Elsevier</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Should endometriosis be an indication for intr...</td>\n",
       "      <td>[http://www.fertstert.org/article/S0015-0282(1...</td>\n",
       "      <td>Fertility and Sterility</td>\n",
       "      <td>104.0</td>\n",
       "      <td>2015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>NaN</td>\n",
       "      <td>[{'name': 'P. M. Voltes'}]</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>[Performance art]</td>\n",
       "      <td>00000081-4562-4882-a29d-5b9cb510ffc3</td>\n",
       "      <td>228.0</td>\n",
       "      <td>[musica, arte]</td>\n",
       "      <td>es</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6</td>\n",
       "      <td>Editorial Lira</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Radio: Radio San Sebastián</td>\n",
       "      <td>[http://dialnet.unirioja.es/servlet/articulo?c...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>20.0</td>\n",
       "      <td>1950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>早期発見と治療成績の向上で担癌患者の生存期間が延びており，それに伴い重複癌を経験する機会も増...</td>\n",
       "      <td>[{'name': '高田和外'}, {'name': 'ほか'}]</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>00000099-1ee9-4741-908c-02a7bf4857e8</td>\n",
       "      <td>4.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>ja</td>\n",
       "      <td>NaN</td>\n",
       "      <td>700</td>\n",
       "      <td>697</td>\n",
       "      <td>癌と化学療法社</td>\n",
       "      <td>NaN</td>\n",
       "      <td>異時性6 重複癌の1 例</td>\n",
       "      <td>[http://www.pieronline.jp/content/article/0385...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>38.0</td>\n",
       "      <td>2011</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                            abstract  \\\n",
       "0  A system and method for maskless direct write ...   \n",
       "1                                                NaN   \n",
       "2                                                NaN   \n",
       "3                                                NaN   \n",
       "4  早期発見と治療成績の向上で担癌患者の生存期間が延びており，それに伴い重複癌を経験する機会も増...   \n",
       "\n",
       "                                             authors doc_type  \\\n",
       "0                                                NaN      NaN   \n",
       "1                   [{'name': 'Ahmed M. Alluwaimi'}]      NaN   \n",
       "2  [{'name': 'Jovana P. Lekovich', 'org': 'Weill ...  Journal   \n",
       "3                         [{'name': 'P. M. Voltes'}]      NaN   \n",
       "4                 [{'name': '高田和外'}, {'name': 'ほか'}]      NaN   \n",
       "\n",
       "                                doi  \\\n",
       "0                               NaN   \n",
       "1          10.4172/2157-7560.C1.030   \n",
       "2  10.1016/j.fertnstert.2015.07.492   \n",
       "3                               NaN   \n",
       "4                               NaN   \n",
       "\n",
       "                                                 fos  \\\n",
       "0  [Electronic engineering, Computer hardware, En...   \n",
       "1      [Biology, Virology, Immunology, Microbiology]   \n",
       "2                                                NaN   \n",
       "3                                  [Performance art]   \n",
       "4                                                NaN   \n",
       "\n",
       "                                     id  issue  \\\n",
       "0  0000002e-c2f2-4e25-9341-60d39130ac7a    NaN   \n",
       "1  0000003d-5ce0-4eed-a56f-bbd3f8d8242b    NaN   \n",
       "2  00000061-0bab-468b-9481-de0c899783f3    3.0   \n",
       "3  00000081-4562-4882-a29d-5b9cb510ffc3  228.0   \n",
       "4  00000099-1ee9-4741-908c-02a7bf4857e8    4.0   \n",
       "\n",
       "                                            keywords lang  n_citation  \\\n",
       "0                                                NaN   en         NaN   \n",
       "1  [paratuberculosis, of, subspecies, proceedings...   en         NaN   \n",
       "2                                                NaN   en         NaN   \n",
       "3                                     [musica, arte]   es         NaN   \n",
       "4                                                NaN   ja         NaN   \n",
       "\n",
       "  page_end page_start            publisher  \\\n",
       "0      NaN        NaN                  NaN   \n",
       "1      NaN        NaN  OMICS International   \n",
       "2      NaN        NaN             Elsevier   \n",
       "3      NaN          6       Editorial Lira   \n",
       "4      700        697              癌と化学療法社   \n",
       "\n",
       "                                          references  \\\n",
       "0  [354c172f-d877-4e60-a7eb-c1b1cf03ce4d, 76cf106...   \n",
       "1                                                NaN   \n",
       "2                                                NaN   \n",
       "3                                                NaN   \n",
       "4                                                NaN   \n",
       "\n",
       "                                               title  \\\n",
       "0  System and Method for Maskless Direct Write Li...   \n",
       "1  The dilemma of the Mycobacterium avium subspec...   \n",
       "2  Should endometriosis be an indication for intr...   \n",
       "3                         Radio: Radio San Sebastián   \n",
       "4                                       異時性6 重複癌の1 例   \n",
       "\n",
       "                                                 url                    venue  \\\n",
       "0  [http://www.freepatentsonline.com/y2016/021111...                      NaN   \n",
       "1  [http://www.omicsonline.org/proceedings/the-di...                      NaN   \n",
       "2  [http://www.fertstert.org/article/S0015-0282(1...  Fertility and Sterility   \n",
       "3  [http://dialnet.unirioja.es/servlet/articulo?c...                      NaN   \n",
       "4  [http://www.pieronline.jp/content/article/0385...                      NaN   \n",
       "\n",
       "   volume  year  \n",
       "0     NaN  2015  \n",
       "1     NaN  2016  \n",
       "2   104.0  2015  \n",
       "3    20.0  1950  \n",
       "4    38.0  2011  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "orig_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(10399, 6)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_df = (orig_df.query('lang == \"en\"')\n",
    "            .drop_duplicates(subset='title', keep='first')\n",
    "            .drop(['doc_type', 'doi', 'id', 'issue', 'lang', 'n_citation',\n",
    "                   'page_end', 'page_start', 'publisher', 'references', 'url', 'venue', 'volume'], axis=1)\n",
    "           )\n",
    "model_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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>abstract</th>\n",
       "      <th>authors</th>\n",
       "      <th>fos</th>\n",
       "      <th>keywords</th>\n",
       "      <th>title</th>\n",
       "      <th>year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A system and method for maskless direct write ...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>[Electronic engineering, Computer hardware, En...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>System and Method for Maskless Direct Write Li...</td>\n",
       "      <td>2015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>NaN</td>\n",
       "      <td>[{'name': 'Ahmed M. Alluwaimi'}]</td>\n",
       "      <td>[Biology, Virology, Immunology, Microbiology]</td>\n",
       "      <td>[paratuberculosis, of, subspecies, proceedings...</td>\n",
       "      <td>The dilemma of the Mycobacterium avium subspec...</td>\n",
       "      <td>2016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NaN</td>\n",
       "      <td>[{'name': 'Jovana P. Lekovich', 'org': 'Weill ...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Should endometriosis be an indication for intr...</td>\n",
       "      <td>2015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>NaN</td>\n",
       "      <td>[{'name': 'George C. Sponsler'}]</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Should APS Discuss Public Issues: Direct, pers...</td>\n",
       "      <td>1968</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Full textFull text is available as a scanned c...</td>\n",
       "      <td>[{'name': 'M. T. Richards'}]</td>\n",
       "      <td>[Medicine, Pathology, Gynecology, Surgery]</td>\n",
       "      <td>[breast neoplasms, female, middle aged, adoles...</td>\n",
       "      <td>Breast surgery--as an office procedure</td>\n",
       "      <td>1973</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                            abstract  \\\n",
       "0  A system and method for maskless direct write ...   \n",
       "1                                                NaN   \n",
       "2                                                NaN   \n",
       "5                                                NaN   \n",
       "7  Full textFull text is available as a scanned c...   \n",
       "\n",
       "                                             authors  \\\n",
       "0                                                NaN   \n",
       "1                   [{'name': 'Ahmed M. Alluwaimi'}]   \n",
       "2  [{'name': 'Jovana P. Lekovich', 'org': 'Weill ...   \n",
       "5                   [{'name': 'George C. Sponsler'}]   \n",
       "7                       [{'name': 'M. T. Richards'}]   \n",
       "\n",
       "                                                 fos  \\\n",
       "0  [Electronic engineering, Computer hardware, En...   \n",
       "1      [Biology, Virology, Immunology, Microbiology]   \n",
       "2                                                NaN   \n",
       "5                                                NaN   \n",
       "7         [Medicine, Pathology, Gynecology, Surgery]   \n",
       "\n",
       "                                            keywords  \\\n",
       "0                                                NaN   \n",
       "1  [paratuberculosis, of, subspecies, proceedings...   \n",
       "2                                                NaN   \n",
       "5                                                NaN   \n",
       "7  [breast neoplasms, female, middle aged, adoles...   \n",
       "\n",
       "                                               title  year  \n",
       "0  System and Method for Maskless Direct Write Li...  2015  \n",
       "1  The dilemma of the Mycobacterium avium subspec...  2016  \n",
       "2  Should endometriosis be an indication for intr...  2015  \n",
       "5  Should APS Discuss Public Issues: Direct, pers...  1968  \n",
       "7             Breast surgery--as an office procedure  1973  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "del orig_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def feature_array(x):\n",
    "    df_list = [pd.DataFrame([[1] * len(val)], columns=val, index=[index])\n",
    "                  if isinstance(val, list) else pd.DataFrame(index=[index])\n",
    "                  for val, index in zip(x.values, x.index)]\n",
    "    feature_df = pd.concat(df_list, axis=1, sort=True)\n",
    "    return feature_df.fillna(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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>1831</th>\n",
       "      <th>1832</th>\n",
       "      <th>1833</th>\n",
       "      <th>1834</th>\n",
       "      <th>1836</th>\n",
       "      <th>1837</th>\n",
       "      <th>1840</th>\n",
       "      <th>1841</th>\n",
       "      <th>1845</th>\n",
       "      <th>1847</th>\n",
       "      <th>...</th>\n",
       "      <th>2008</th>\n",
       "      <th>2009</th>\n",
       "      <th>2010</th>\n",
       "      <th>2011</th>\n",
       "      <th>2012</th>\n",
       "      <th>2013</th>\n",
       "      <th>2014</th>\n",
       "      <th>2015</th>\n",
       "      <th>2016</th>\n",
       "      <th>2017</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2 rows × 156 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   1831  1832  1833  1834  1836  1837  1840  1841  1845  1847  ...  2008  \\\n",
       "0     0     0     0     0     0     0     0     0     0     0  ...     0   \n",
       "1     0     0     0     0     0     0     0     0     0     0  ...     0   \n",
       "\n",
       "   2009  2010  2011  2012  2013  2014  2015  2016  2017  \n",
       "0     0     0     0     0     0     0     1     0     0  \n",
       "1     0     0     0     0     0     0     0     1     0  \n",
       "\n",
       "[2 rows x 156 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "year_features = pd.get_dummies(model_df['year'].astype('category'))\n",
    "year_features.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Size of first feature array:  3787893471\n"
     ]
    }
   ],
   "source": [
    "fos_features = feature_array(model_df['fos'])\n",
    "first_features = fos_features.join(year_features).T\n",
    "\n",
    "from sys import getsizeof\n",
    "print('Size of first feature array: ', getsizeof(first_features))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "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>5</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>...</th>\n",
       "      <th>19985</th>\n",
       "      <th>19986</th>\n",
       "      <th>19987</th>\n",
       "      <th>19988</th>\n",
       "      <th>19993</th>\n",
       "      <th>19994</th>\n",
       "      <th>19995</th>\n",
       "      <th>19997</th>\n",
       "      <th>19998</th>\n",
       "      <th>19999</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Electronic engineering</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Computer hardware</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Engineering</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Engineering drawing</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Biology</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 10399 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                        0      1      2      5      7      8      9      \\\n",
       "Electronic engineering    1.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
       "Computer hardware         1.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
       "Engineering               1.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
       "Engineering drawing       1.0    0.0    0.0    0.0    0.0    0.0    0.0   \n",
       "Biology                   0.0    1.0    0.0    0.0    0.0    0.0    0.0   \n",
       "\n",
       "                        10     11     12     ...  19985  19986  19987  19988  \\\n",
       "Electronic engineering    0.0    0.0    0.0  ...    0.0    0.0    0.0    0.0   \n",
       "Computer hardware         0.0    0.0    0.0  ...    0.0    0.0    0.0    0.0   \n",
       "Engineering               0.0    0.0    0.0  ...    0.0    0.0    0.0    0.0   \n",
       "Engineering drawing       0.0    0.0    0.0  ...    0.0    0.0    0.0    0.0   \n",
       "Biology                   0.0    0.0    0.0  ...    0.0    0.0    0.0    0.0   \n",
       "\n",
       "                        19993  19994  19995  19997  19998  19999  \n",
       "Electronic engineering    0.0    0.0    0.0    0.0    0.0    0.0  \n",
       "Computer hardware         0.0    0.0    0.0    0.0    0.0    0.0  \n",
       "Engineering               0.0    0.0    0.0    0.0    0.0    0.0  \n",
       "Engineering drawing       0.0    0.0    0.0    0.0    0.0    0.0  \n",
       "Biology                   0.0    0.0    0.0    0.0    0.0    0.0  \n",
       "\n",
       "[5 rows x 10399 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "first_features.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy.spatial.distance import cosine\n",
    "\n",
    "def item_collab_filter(features_df):\n",
    "    return pd.DataFrame([[1 - cosine(col_val1, col_val2)\n",
    "                                     for col_val1 in features_df.T.values] for col_val2 in features_df.T.values],\n",
    "                                     index = features_df.columns, columns = features_df.columns)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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>5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.447214</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.447214</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0    1         2    5\n",
       "0  1.000000  0.0  0.447214  0.0\n",
       "1  0.000000  1.0  0.000000  0.0\n",
       "2  0.447214  0.0  1.000000  0.0\n",
       "5  0.000000  0.0  0.000000  1.0"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "item_collab_filter(first_features.loc[:, 0:5])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "first_items = item_collab_filter(first_features.loc[:, 0:1000])\n",
    "first_items.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sns\n",
    "import numpy as np\n",
    "\n",
    "sns.set()\n",
    "ax = sns.heatmap(first_items.fillna(0), \n",
    "                 vmin=0, vmax=1, \n",
    "                 cmap='YlGnBu', \n",
    "                 xticklabels=250, yticklabels=250)\n",
    "ax.tick_params(labelsize=12)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Year spread: 1831 - 2017\n",
      "Quantile spread:\n",
      " 0.25    1990.0\n",
      "0.50    2005.0\n",
      "0.75    2012.0\n",
      "Name: year, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "year_min = model_df['year'].min()\n",
    "year_max = model_df['year'].max()\n",
    "print('Year spread:', year_min, '-', year_max)\n",
    "print('Quantile spread:\\n', model_df['year'].quantile([0.25, 0.5, 0.75]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f905f8bcd30>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 500x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "model_df['year'].hist(bins=year_max - year_min)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "18"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bins = int(round(year_max - year_min) / 10)\n",
    "bins"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "temp_df = pd.DataFrame(index=model_df.index)\n",
    "temp_df['yearBinned'] = pd.cut(model_df['year'].tolist(), bins, precision=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "IntervalIndex([(1831.0, 1841.0], (1841.0, 1852.0], (1852.0, 1862.0], (1862.0, 1872.0], (1872.0, 1883.0] ... (1965.0, 1976.0], (1976.0, 1986.0], (1986.0, 1996.0], (1996.0, 2007.0], (2007.0, 2017.0]],\n",
       "              closed='right',\n",
       "              dtype='interval[float64]')"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_yrs = pd.get_dummies(temp_df['yearBinned'])\n",
    "X_yrs.columns.categories"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f905f7ea2b0>"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 500x300 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "X_yrs.sum().plot(kind='bar')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Our pandas Series, in bytes:  3772086600\n",
      "Our hashed numpy array, in bytes:  112\n"
     ]
    }
   ],
   "source": [
    "X_fos = fos_features.values\n",
    "\n",
    "print('Our pandas Series, in bytes: ', getsizeof(fos_features))\n",
    "print('Our hashed numpy array, in bytes: ', getsizeof(X_fos))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(10399, 45337)"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_fos.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3771675704"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_fos.nbytes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The power of feature engineering saves us, in bytes:  -3773173256\n"
     ]
    }
   ],
   "source": [
    "second_features = np.append(X_fos, X_yrs, axis = 1)\n",
    "\n",
    "print('The power of feature engineering saves us, in bytes: ',\n",
    "         getsizeof(first_features) - getsizeof(second_features))\n",
    "\n",
    "from sklearn.metrics.pairwise import cosine_similarity\n",
    "\n",
    "def piped_collab_filter(features_matrix, index, top_n):\n",
    "    item_similarities = \\\n",
    "        1 - cosine_similarity(features_matrix[index:index+1],\n",
    "                              features_matrix).flatten()\n",
    "    related_indices = \\\n",
    "        [i for i in item_similarities.argsort()[::-1] if i != index]\n",
    "    return [(index, item_similarities[index]) for index in related_indices][0:top_n]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3785.0176"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "first_features.memory_usage().sum()/1000/1000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3773.1731600000003"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "second_features.nbytes/1000/1000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The power of feature engineering saves us, in bytes:  11844440\n"
     ]
    }
   ],
   "source": [
    "print('The power of feature engineering saves us, in bytes: ',\n",
    "         first_features.memory_usage().sum() - second_features.nbytes)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "filled_df = model_df.fillna('None')\n",
    "\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "\n",
    "vectorizer = TfidfVectorizer(sublinear_tf=True, max_df=0.5, stop_words='english')\n",
    "X_abstract = vectorizer.fit_transform(filled_df['abstract'])\n",
    "\n",
    "third_features = np.append(second_features, X_abstract.toarray(), axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[{'None': 0},\n",
       " {'Ahmed M. Alluwaimi': 1},\n",
       " {'Jovana P. Lekovich': 2, 'Weill Cornell Medical College, New York, NY': 2},\n",
       " {'George C. Sponsler': 5},\n",
       " {'M. T. Richards': 7}]"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "authors_list = []\n",
    "\n",
    "for row in filled_df.itertuples():\n",
    "    # それぞれのSeriesインデックスからdict型オブジェクトを作成する\n",
    "    if isinstance(row.authors, str):\n",
    "        y = {'None': row.Index}\n",
    "    if isinstance(row.authors, list):\n",
    "        # これらのキー、値をdict型オブジェクトに追加する\n",
    "        y = dict.fromkeys(row.authors[0].values(), row.Index)\n",
    "    authors_list.append(y)\n",
    "\n",
    "authors_list[0:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.feature_extraction import DictVectorizer\n",
    "v = DictVectorizer(sparse=False)\n",
    "D = authors_list\n",
    "X_authors = v.fit_transform(D)\n",
    "fourth_features = np.append(third_features, X_authors, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "10399"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(D)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(10399, 14028)"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_authors.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "del second_features\n",
    "del third_features\n",
    "del X_fos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Based on the paper:\n",
      "Paper index =  2\n",
      "Title : Should endometriosis be an indication for intracytoplasmic sperm injection (ICSI) in fresh IVF cycles\n",
      "FOS : nan\n",
      "Year : 2015\n",
      "Abstract : nan\n",
      "Authors : [{'name': 'Jovana P. Lekovich', 'org': 'Weill Cornell Medical College, New York, NY'}, {'name': 'G.D. Palermo', 'org': 'Weill Medical College of Cornell University, New York, NY'}, {'name': 'Nigel Pereira', 'org': 'The Ronald O. Perelman and Claudia Cohen Center, New York, NY'}, {'name': 'Zev Rosenwaks', 'org': 'Weill Cornell Medical College, New York, NY'}] \n",
      "\n"
     ]
    }
   ],
   "source": [
    "def paper_recommender(items_df, paper_ix, top_n):\n",
    "    if paper_ix in model_df.index:\n",
    "        print('Based on the paper:')\n",
    "        print('Paper index = ', model_df.loc[paper_ix].name)\n",
    "        print('Title :', model_df.loc[paper_ix]['title'])\n",
    "        print('FOS :', model_df.loc[paper_ix]['fos'])\n",
    "        print('Year :', model_df.loc[paper_ix]['year'])\n",
    "        print('Abstract :', model_df.loc[paper_ix]['abstract'])\n",
    "        print('Authors :', model_df.loc[paper_ix]['authors'], '\\n')\n",
    "\n",
    "        array_ix = model_df.index.get_loc(paper_ix)\n",
    "        top_results = piped_collab_filter(items_df, array_ix, top_n)\n",
    "        print('\\nTop',top_n,'results: ')\n",
    "\n",
    "        order = 1\n",
    "        for i in range(len(top_results)):\n",
    "            print(order,'. Paper index = ', model_df.iloc[top_results[i][0]].name)\n",
    "            print('Similarity score: ', top_results[i][1])\n",
    "            print('Title :', model_df.iloc[top_results[i][0]]['title'])\n",
    "            print('FOS :', model_df.iloc[top_results[i][0]]['fos'])\n",
    "            print('Year :', model_df.iloc[top_results[i][0]]['year'])\n",
    "            print('Abstract :', model_df.iloc[top_results[i][0]]['abstract'])\n",
    "            print('Authors :', model_df.iloc[top_results[i][0]]['authors'], '\\n')\n",
    "            if order < top_n: order += 1\n",
    "                    \n",
    "    else:\n",
    "        print('Whoops! Choose another paper. Try something from here: \\n', model_df.index[100:200])\n",
    "\n",
    "paper_recommender(fourth_features, 2, 3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}