KLabNameDisBeta.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/weight_boosting.py:29: DeprecationWarning: numpy.core.umath_tests is an internal NumPy module and should not be imported. It will be removed in a future NumPy release.\n",
      "  from numpy.core.umath_tests import inner1d\n"
     ]
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "import sys\n",
    "import random\n",
    "import pickle\n",
    "import pylab as plt\n",
    "import networkx as nx\n",
    "import time\n",
    "import numpy as np\n",
    "import itertools\n",
    "from itertools import combinations\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.ensemble import GradientBoostingClassifier\n",
    "from xgboost import XGBClassifier\n",
    "from collections import defaultdict, Counter\n",
    "path='/Users/lingfeiw/Documents/research/teamscience/team/BillNameDis/'\n",
    "\n",
    "# flush print\n",
    "def flushPrint(d):\n",
    "    sys.stdout.write('\\r')\n",
    "    sys.stdout.write('%s' % d)\n",
    "    sys.stdout.flush()\n",
    "    \n",
    "def make_meshgrid(x, y, h=.02):\n",
    "    x_min, x_max = x.min() - 1, x.max() + 1\n",
    "    y_min, y_max = y.min() - 1, y.max() + 1\n",
    "    xx, yy = np.meshgrid(np.arange(x_min, x_max, h),\n",
    "                         np.arange(y_min, y_max, h))\n",
    "    return xx, yy\n",
    "\n",
    "# similarity functions\n",
    "#def sa(i,j): return len(set(A[i]).intersection(A[j]))/float(min(len(A[i]),len(A[j])))\n",
    "def sa(i,j): return (len(set(A[i]).intersection(A[j]))-1)/(min(len(A[i]),len(A[j]))-1)# remove the author that tie these two papers together\n",
    "def sr(i,j): return len(set(R[i]).intersection(R[j]))/min(len(R[i]),len(R[j]))\n",
    "def sc(i,j): return len(set(C[i]).intersection(C[j]))/min(len(C[i]),len(C[j]))\n",
    "def sx(i,j): return int(i in R[j])+int(j in R[i])\n",
    "\n",
    "def disambiguous(papers):\n",
    "    # iterate over all possible pairs and construct data\n",
    "    X=np.array([[sa(i,j),sr(i,j),sc(i,j),sx(i,j),i,j] for i,j in itertools.combinations(sorted(papers), 2)])\n",
    "    # predict linking probability\n",
    "    y=clt.predict(X[:,:-2])\n",
    "    # connect nodes\n",
    "    G = nx.Graph()\n",
    "    G.add_nodes_from(papers)\n",
    "    G.add_edges_from(X[:,-2:][y==1].astype(int))\n",
    "    # export data of component\n",
    "    data = dict((n,c) for n,c in enumerate(sorted(nx.connected_components(G), key=len, reverse=True)))\n",
    "    return data\n",
    "\n",
    "path='/Users/lingfeiw/Documents/research/teamscience/team/BillNameDis/'\n",
    "name='AllEmailsTrainedGradientBoostingClassifier.sav'\n",
    "clt=pickle.load(open(path+name, 'rb'))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Citation data:\n",
    "\n",
    "cited_ID citing_ID1, citing_ID2, ....\n",
    "\n",
    "Reference data:\n",
    "\n",
    "citing_ID cited_ID1, cited_ID2, ....\n",
    "\n",
    "Author data:\n",
    "\n",
    "paper_id, wos_standard_name1, wos_standard_name2, ...\n",
    "\n",
    "effectivePapers\n",
    "\n",
    "A & C & R"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "221.0"
     ]
    }
   ],
   "source": [
    "'''\n",
    "# effective papers\n",
    "P=set([])\n",
    "n=0\n",
    "with open('/Users/lingfeiw/Documents/research/teamscience/team/BillNameDis/effectivePapers.txt', \"r\") as f:\n",
    "    for line in f:\n",
    "        n+=1\n",
    "        if n%100000==0:\n",
    "            flushPrint(n/100000)#221\n",
    "        line=line.strip().split('\\t')   \n",
    "        P.add(int(line[0]))\n",
    "        \n",
    "# citation\n",
    "C={}\n",
    "n=0\n",
    "with open('/Users/lingfeiw/Documents/research/teamscience/team/BillNameDis/PaperCitationList.txt','r') as f:\n",
    "    for line in f:\n",
    "        n+=1\n",
    "        if n%100000==0:\n",
    "            flushPrint(n/100000)#286\n",
    "        line=list(map(int,line.strip().split('\\t')))\n",
    "        p=int(line[0])\n",
    "        if p in P:\n",
    "            C[p] = line[1:]\n",
    "            \n",
    "# reference \n",
    "R={}\n",
    "n=0\n",
    "with open('/Users/lingfeiw/Documents/research/teamscience/team/reference1900_2015.txt','r') as f:\n",
    "    for line in f:\n",
    "        n+=1\n",
    "        if n%100000==0:\n",
    "            flushPrint(n/100000)#314 \n",
    "        line=list(map(int,line.strip().split('\\t')))\n",
    "        p=line[0]\n",
    "        if p in P:\n",
    "            R[p] = line[1:]\n",
    "\n",
    "#author name\n",
    "A={}\n",
    "n=0\n",
    "with open('/Users/lingfeiw/Documents/research/teamscience/team/authorlist1900_2015.txt','r') as f:\n",
    "    for line in f:\n",
    "        n+=1\n",
    "        if n%100000==0:\n",
    "            flushPrint(n/100000)#427 \n",
    "        line=line.strip().split('\\t')\n",
    "        p=int(line[0])\n",
    "        if p in P:\n",
    "            A[p] = [i.lower() for i in line[1:]]\n",
    "            \n",
    "pickle.dump(C, open(path+'citation.pkl', 'wb'), -1)\n",
    "pickle.dump(R, open(path+'reference.pkl', 'wb'), -1)\n",
    "pickle.dump(A, open(path+'author.pkl', 'wb'), -1)\n",
    "pickle.dump(P, open(path+'effectivePapers.pkl', 'wb'), -1)\n",
    "'''\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# ~35 minutes\n",
    "with open(path+'author.pkl', 'rb') as f:\n",
    "    A = pickle.load(f) \n",
    "with open(path+'citation.pkl', 'rb') as f:\n",
    "    C = pickle.load(f) \n",
    "with open(path+'reference.pkl', 'rb') as f:\n",
    "    R = pickle.load(f) \n",
    "with open(path+'effectivePapers.pkl', 'rb') as f:\n",
    "    P = pickle.load(f)\n",
    "    \n",
    "#len(R),len(A),len(C) #22150464"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(22150464, 22150464, 22150464)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(R),len(A),len(C)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Email data training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "153.0"
     ]
    }
   ],
   "source": [
    "E=defaultdict(lambda:{})\n",
    "n=0\n",
    "with open('/Users/lingfeiw/Documents/research/teamscience/team/rawAuthorEmailPaperWOS.txt', \"r\") as f:\n",
    "    for line in f:\n",
    "        n+=1\n",
    "        if n%100000==0:\n",
    "            flushPrint(n/100000)#153\n",
    "        line=line.strip().split('\\t')\n",
    "        author,email=line[0].split('+',1)\n",
    "        author=author.lower()\n",
    "        if email!='?':\n",
    "            vs=[int(i) for i in line[1:] if int(i) in P]\n",
    "            if len(vs)>=2:# with real email and two or more papers relevant to that email\n",
    "                E[author][email]=vs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "867895"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(E)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "867.0"
     ]
    }
   ],
   "source": [
    "Sim=[]\n",
    "Label=[]\n",
    "n=0\n",
    "for e in E:\n",
    "    n+=1\n",
    "    if n%1000==0:\n",
    "        flushPrint(n/1000) #868\n",
    "    # merge all papers\n",
    "    papers=sorted([j for i in E[e] for j in E[e][i]])\n",
    "    # construct dictionary to compare the author-emails of a pair of papers \n",
    "    es=dict((j,i) for i in E[e] for j in E[e][i])\n",
    "    if 2<=len(E[e])<=10 and len(papers)<=200:\n",
    "        for i,j in itertools.combinations(sorted(papers), 2):\n",
    "            Sim.append([sa(i,j),sr(i,j),sc(i,j),sx(i,j)])\n",
    "            Label.append(int(es[i]==es[j]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(Counter({1: 30278129, 0: 31081417}), 61359546)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Counter(Label),len(Label)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "239.30703496932983\n"
     ]
    }
   ],
   "source": [
    "st=time.time()\n",
    "forest = GradientBoostingClassifier(n_estimators=3, learning_rate=1.0).fit(Sim, Label)\n",
    "print(time.time()-st)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6703493373304946"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "forest.score(Sim, Label)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "#path='/Users/lingfeiw/Documents/research/teamscience/team/BillNameDis/'\n",
    "#name='AllEmailsTrainedGradientBoostingClassifier.sav'\n",
    "#pickle.dump(forest, open(path+name, 'wb'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 'EVANS, JA' testing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "153.0"
     ]
    }
   ],
   "source": [
    "E={}\n",
    "n=0\n",
    "with open('/Users/lingfeiw/Documents/research/teamscience/team/rawAuthorEmailPaperWOS.txt', \"r\") as f:\n",
    "    for line in f:\n",
    "        n+=1\n",
    "        if n%100000==0:\n",
    "            flushPrint(n/100000)#153\n",
    "        line=line.strip().split('\\t')\n",
    "        author,email=line[0].split('+',1)\n",
    "        if author.lower()=='evans, ja' and email!='?':\n",
    "            vs=[int(i) for i in line[1:] if int(i) in P]\n",
    "            if len(vs)>=2:# with real email and two or more papers relevant to that email\n",
    "                E[line[0]]=vs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "metadata": {},
   "outputs": [],
   "source": [
    "papers=set([j for k in E for j in E[k]])\n",
    "data=disambiguous(papers)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 183,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "42"
      ]
     },
     "execution_count": 183,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(papers)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "metadata": {},
   "outputs": [],
   "source": [
    "# visualization \n",
    "X=np.array([[sa(i,j),sr(i,j),sc(i,j),sx(i,j),i,j] for i,j in itertools.combinations(sorted(papers), 2)])\n",
    "y=clt.predict(X[:,:-2])\n",
    "G = nx.Graph()\n",
    "G.add_nodes_from(papers)\n",
    "G.add_edges_from(X[:,-2:][y==1].astype(int))\n",
    "E_={j:list(E.keys()).index(k) for k in E for j in E[k]}\n",
    "data_={j:k for k in data for j in data[k]}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(42, 109, 11, 10)"
      ]
     },
     "execution_count": 185,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(G.nodes()),len(G.edges()),len(data),len(E)#(38, 109, 7,10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 204,
   "metadata": {},
   "outputs": [],
   "source": [
    "pos = nx.spring_layout(G,k=0.2,iterations=30)\n",
    "x,y,emailGroup,simGroup=np.array([list(pos[i])+[E_[i],data_[i]] for i in pos]).T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 242,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig=plt.figure(figsize=(8, 6),facecolor='white')\n",
    "ax = fig.add_subplot(111)\n",
    "nx.draw_networkx_edges(G,pos,alpha=0.3,width=0.5,arrows=False,zorder=1)\n",
    "plt.scatter(x,y,s=50,c=emailGroup,cmap='tab10',edgecolor='gray',alpha=1,zorder=2)\n",
    "plt.colorbar()\n",
    "plt.xlim(-1,1)\n",
    "plt.ylim(-1,1)\n",
    "ax.grid(linestyle='-', linewidth='0.5', color='orange',zorder=0)\n",
    "plt.title('EVANS, JA',size=14)\n",
    "#plt.savefig('/Users/lingfeiw/Documents/research/teamscience/team/BillNameDis/evansJAnewMarked.pdf')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 243,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['EVANS, JA+jaevans@ucsd.edu', 'EVANS, JA+je@bgs.ac.uk', 'EVANS, JA+j.a.evans@bristol.ac.uk', 'EVANS, JA+evans@ices.utexas.edu', 'EVANS, JA+je@nigl.nerc.ac.uk', 'EVANS, JA+evans@kccr.de', 'EVANS, JA+jevans@ms.umanitoba.ca', 'EVANS, JA+jevans@uchicago.edu', 'EVANS, JA+jaredaevans@gmail.com', 'EVANS, JA+jevans@msm.edu'])"
      ]
     },
     "execution_count": 243,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "E.keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('EVANS, JA+je@bgs.ac.uk', 'EVANS, JA+je@nigl.nerc.ac.uk')"
      ]
     },
     "execution_count": 121,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#list(sorted(zip(simGroup,emailGroup)))\n",
    "#1,4 and 0,9\n",
    "list(E.keys())[1],list(E.keys())[4] #https://www.bgs.ac.uk/staff/profiles/0394.html  school of Archaeology"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('EVANS, JA+jaevans@ucsd.edu', 'EVANS, JA+jevans@msm.edu')"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "list(E.keys())[0],list(E.keys())[9] # department of psychology"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Ground truth testing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\r",
      "1.0"
     ]
    }
   ],
   "source": [
    "'''\n",
    "O={}\n",
    "n=0\n",
    "with open('/Users/lingfeiw/Documents/research/teamscience/team/orcid/ORCIDmatchedScholars_byTitleDoi.txt','r') as f:\n",
    "    for line in f:\n",
    "        n+=1\n",
    "        if n%100000==0:\n",
    "            flushPrint(n/100000)\n",
    "        line=line.strip().split('\\t')\n",
    "        O[int(line[0])]=list(map(int,line[1:]))\n",
    "        \n",
    "len(O)#155419\n",
    "\n",
    "AU={}\n",
    "n=0\n",
    "with open('/Users/lingfeiw/Documents/research/teamscience/team/authorlist1900_2015.txt','r') as f:\n",
    "    for line in f:\n",
    "        n+=1\n",
    "        if n%100000==0:\n",
    "            flushPrint(n/100000)#427\n",
    "        line=line.strip().split('\\t')\n",
    "        p=int(line[0])\n",
    "        AU[p]=line[1:]\n",
    "        \n",
    "def matchAuthor(name):\n",
    "    K=defaultdict(lambda:0)\n",
    "    m=0\n",
    "    for p in O[name]:\n",
    "        if p in AU:\n",
    "            m+=1\n",
    "            for n in AU[p]:\n",
    "                K[n]+=1\n",
    "    vs=[k for k,v in K.items() if v==m]\n",
    "    if len(vs)==1:\n",
    "        return vs[0]\n",
    "    \n",
    "M=[]\n",
    "n=0\n",
    "for name in O:\n",
    "    n+=1\n",
    "    if n%1000==0:\n",
    "        flushPrint(n/1000)#150\n",
    "    au=matchAuthor(name)\n",
    "    if au:\n",
    "        M.append([au.lower()]+O[name])\n",
    "\n",
    "with open(path+'ORCIDwithAuthorName.txt', 'w') as f:\n",
    "    for v in M:\n",
    "        f.write('\\t'.join(map(str,v))+'\\n')\n",
    "'''"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "N=[]\n",
    "with open(path+'ORCIDwithAuthorName.txt', 'r') as f:\n",
    "    for line in f:\n",
    "        line=line.strip().split('\\t')\n",
    "        N.append([line[0]]+list(map(int,line[1:])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "names=set([i[0] for i in N])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(46642, 50189)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(names),len(N)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#nordquist, n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "82.064\n"
     ]
    }
   ],
   "source": [
    "n=0\n",
    "with open('/Users/lingfeiw/Documents/research/teamscience/team/BillNameDis/PaperListEffective.txt', \"r\") as f:\n",
    "    for line in f:\n",
    "        n+=1\n",
    "        if n%100000==0:\n",
    "            flushPrint(n/100000)#84\n",
    "        line=line.strip().split('\\t')\n",
    "        author=line[0].lower()   # first break down               # second break down\n",
    "        if author =='yeh, tf':# #'nordquist, n' #n==7801155 #     #'yeh, tf' #n==8207337 #\n",
    "            print(len(line)-1)\n",
    "            break"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "8207337"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "826.0"
     ]
    }
   ],
   "source": [
    "# export predictions \n",
    "with open(path+'ORCIDprediction3.txt', 'w') as f1:\n",
    "    n=0\n",
    "    with open('/Users/lingfeiw/Documents/research/teamscience/team/BillNameDis/PaperListEffective.txt', \"r\") as f:\n",
    "        for line in f:\n",
    "            n+=1\n",
    "            #if n>7801155:\n",
    "            if n>8207337:\n",
    "                if n%10000==0:\n",
    "                    flushPrint(n/10000)#840\n",
    "                line=line.strip().split('\\t')\n",
    "                author=line[0].lower()\n",
    "                if author in names:\n",
    "                    papers=list(map(int,line[1:]))\n",
    "                    if len(papers)<=100:\n",
    "                            dic=disambiguous(papers) \n",
    "                            s='\\t'.join(['__'.join((str(k),'_'.join(map(str,sorted(v))))) for k,v in dic.items()])\n",
    "                            f1.write(author+'\\t'+s+'\\n')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "84.0"
     ]
    }
   ],
   "source": [
    "n=0\n",
    "PP=[]\n",
    "with open('/Users/lingfeiw/Documents/research/teamscience/team/BillNameDis/PaperListEffective.txt', \"r\") as f:\n",
    "    for line in f:\n",
    "        n+=1\n",
    "        if n%100000==0:\n",
    "            flushPrint(n/100000)#84\n",
    "        if n>7801155:\n",
    "            line=line.strip().split('\\t')\n",
    "            author=line[0].lower()\n",
    "            if author in names:\n",
    "                papers=list(map(int,line[1:]))\n",
    "                PP.append([author,len(papers)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "79.0"
     ]
    }
   ],
   "source": [
    "# export predictions \n",
    "with open(path+'ORCIDprediction.txt', 'w') as f1:\n",
    "    n=0\n",
    "    with open('/Users/lingfeiw/Documents/research/teamscience/team/BillNameDis/PaperListEffective.txt', \"r\") as f:\n",
    "        for line in f:\n",
    "            n+=1\n",
    "            if n%100000==0:\n",
    "                flushPrint(n/100000)#84\n",
    "            line=line.strip().split('\\t')\n",
    "            author=line[0].lower()\n",
    "            if author in names:\n",
    "                papers=list(map(int,line[1:]))\n",
    "                if len(papers)==1:\n",
    "                    dic={0:set(papers)}\n",
    "                else:\n",
    "                    dic=disambiguous(papers) \n",
    "                s='\\t'.join(['__'.join((str(k),'_'.join(map(str,sorted(v))))) for k,v in dic.items()])\n",
    "                f1.write(author+'\\t'+s+'\\n')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import predictions\n",
    "U={}\n",
    "with open(path+'ORCIDprediction.txt', 'r') as f:\n",
    "    for line in f:\n",
    "        line=line.strip().split('\\t')\n",
    "        author=line[0]\n",
    "        papers=[i.split('__') for i in line[1:]] \n",
    "        dic={int(k):list(map(int,v.split('_'))) \\\n",
    "             for k,v in papers}\n",
    "        U[author]=dic\n",
    "with open(path+'ORCIDprediction2.txt', 'r') as f:\n",
    "    for line in f:\n",
    "        line=line.strip().split('\\t')\n",
    "        author=line[0]\n",
    "        papers=[i.split('__') for i in line[1:]] \n",
    "        dic={int(k):list(map(int,v.split('_'))) \\\n",
    "             for k,v in papers}\n",
    "        U[author]=dic\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "40346"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(U)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "50189"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(N)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "i=N[0]\n",
    "name=i[0]\n",
    "\n",
    "data=[]\n",
    "num=0\n",
    "l2=len(set(U[name][num]).intersection(papers))\n",
    "l2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "50.0"
     ]
    }
   ],
   "source": [
    "dd=[]\n",
    "n=0\n",
    "for i in N:\n",
    "    n+=1\n",
    "    if n%1000==0:\n",
    "        flushPrint(n/1000)#45\n",
    "    name=i[0]\n",
    "    if name in U:\n",
    "        papers=set(i[1:])\n",
    "        if len(papers)>=3:\n",
    "            l0=len(papers)\n",
    "            data=[]\n",
    "            for num in U[name]:\n",
    "                l2=len(set(U[name][num]).intersection(papers))\n",
    "                l1=len(U[name][num])\n",
    "                precision=l2/float(l1)\n",
    "                recall=l2/float(l0)\n",
    "                data.append([recall,precision,l1])\n",
    "            dd.append(sorted(data)[-1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "37620"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(dd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[[0.1, 1.0, 1],\n",
       " [0.6666666666666666, 0.6956521739130435, 23],\n",
       " [0.6666666666666666, 1.0, 6]]"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dd[:3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0.48170425,  0.69676335, 10.9800363 ])"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(dd,axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0.50537871,  0.72492725, 11.29819245])"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.mean(dd,axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.5, 1. , 7. ])"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.median(dd,axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.58333333, 1.        , 7.        ])"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.median(dd,axis=0)"
   ]
  },
  {
   "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": "3.6.8"
  },
  "toc": {
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}