{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "df5a1f59",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from collections import Counter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "0b47ecac",
   "metadata": {},
   "outputs": [],
   "source": [
    "author = pd.read_csv('../data/ht_class/ht_cleaned_author_df.csv')\n",
    "paper = pd.read_csv('../data/ht_class/ht_cleaned_paper_df.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "d8d8d536",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(52, 27)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# first conference papers\n",
    "first_conf_papers = paper[paper.Year == min(paper.Year)]\n",
    "first_conf_papers.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "a857697a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "118"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# first conf authors\n",
    "first_conf_authors = author[author.Year == min(author.Year)]\n",
    "unique_first_conf_authors = list(\n",
    "    set(first_conf_authors['OpenAlex Author ID']))\n",
    "len(unique_first_conf_authors)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "4fead9a0",
   "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>Year</th>\n",
       "      <th>DOI</th>\n",
       "      <th>Title</th>\n",
       "      <th>Number of Authors</th>\n",
       "      <th>Author Position</th>\n",
       "      <th>Author Name</th>\n",
       "      <th>OpenAlex Author ID</th>\n",
       "      <th>Affiliation Name</th>\n",
       "      <th>Affiliation Country Code</th>\n",
       "      <th>Affiliation Type</th>\n",
       "      <th>Binary Type</th>\n",
       "      <th>Cross-type Collaboration</th>\n",
       "      <th>International Collaboration</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7746</th>\n",
       "      <td>1990</td>\n",
       "      <td>10.1109/VISUAL.1990.146385</td>\n",
       "      <td>A journey into the fourth dimension</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>E.S. Panduranga</td>\n",
       "      <td>https://openalex.org/A2302757246</td>\n",
       "      <td>department of computer sciences johns hopkins ...</td>\n",
       "      <td>US</td>\n",
       "      <td>education</td>\n",
       "      <td>education</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9955</th>\n",
       "      <td>1990</td>\n",
       "      <td>10.1109/VISUAL.1990.146393</td>\n",
       "      <td>A numerical method for rendering spherical ref...</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>E.S. Panduranga</td>\n",
       "      <td>https://openalex.org/A2302757246</td>\n",
       "      <td>department of computer sciences johns hopkins ...</td>\n",
       "      <td>US</td>\n",
       "      <td>education</td>\n",
       "      <td>education</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Year                         DOI  \\\n",
       "7746  1990  10.1109/VISUAL.1990.146385   \n",
       "9955  1990  10.1109/VISUAL.1990.146393   \n",
       "\n",
       "                                                  Title  Number of Authors  \\\n",
       "7746                A journey into the fourth dimension                2.0   \n",
       "9955  A numerical method for rendering spherical ref...                3.0   \n",
       "\n",
       "      Author Position      Author Name                OpenAlex Author ID  \\\n",
       "7746              2.0  E.S. Panduranga  https://openalex.org/A2302757246   \n",
       "9955              2.0  E.S. Panduranga  https://openalex.org/A2302757246   \n",
       "\n",
       "                                       Affiliation Name  \\\n",
       "7746  department of computer sciences johns hopkins ...   \n",
       "9955  department of computer sciences johns hopkins ...   \n",
       "\n",
       "     Affiliation Country Code Affiliation Type Binary Type  \\\n",
       "7746                       US        education   education   \n",
       "9955                       US        education   education   \n",
       "\n",
       "      Cross-type Collaboration  International Collaboration  \n",
       "7746                     False                         True  \n",
       "9955                     False                        False  "
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# this person appeared twice in year 1990\n",
    "dic = dict(Counter(first_conf_authors['OpenAlex Author ID']))\n",
    "twice_id = [x for x in dic.keys() if dic[x] > 1]\n",
    "first_conf_authors[first_conf_authors['OpenAlex Author ID'] == twice_id[0]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "46813419",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Counter({'US': 109, 'DE': 4, 'AU': 3, 'CA': 1, 'FR': 2})"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# first conf author countries\n",
    "Counter(first_conf_authors['Affiliation Country Code'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "e3fd43f2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "117"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# how many rows miss openalex author id\n",
    "author[author['OpenAlex Author ID'].isnull()].shape[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "c70319a5",
   "metadata": {},
   "outputs": [],
   "source": [
    "# total number of unique authors\n",
    "author_id_nonan = author[author['OpenAlex Author ID'].notnull()]\n",
    "unique_total_authors = list(set(author_id_nonan['OpenAlex Author ID']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "c4944f5b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "6299"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(unique_total_authors)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "87ce81d1",
   "metadata": {},
   "outputs": [],
   "source": [
    "# total number of unique countries\n",
    "unique_cntry_total_authors = list(set(author_id_nonan['Affiliation Country Code']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "2e8d450d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "42"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(unique_cntry_total_authors)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ccff5bc3",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}