10-doc2vec.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Doc2Vec\n",
    "\n",
    "Le and Mikolov in 2014 introduced the Doc2Vec algorithm, which usually outperforms such simple-averaging of Word2Vec vectors.\n",
    "\n",
    "https://cs.stanford.edu/~quocle/paragraph_vector.pdf\n",
    "\n",
    "The basic idea is: act as if a document has another floating word-like vector, which contributes to all training predictions, and is updated like other word-vectors, but we will call it a doc-vector. Gensim’s Doc2Vec class implements this algorithm.\n",
    "\n",
    "\n",
    "https://radimrehurek.com/gensim/models/doc2vec.html\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-07-04T07:31:51.371380Z",
     "start_time": "2021-07-04T07:31:50.589389Z"
    }
   },
   "outputs": [],
   "source": [
    "from gensim.test.utils import common_texts\n",
    "from gensim.models.doc2vec import Doc2Vec, TaggedDocument\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "plt.style.use('ggplot')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-07-05T06:52:53.336835Z",
     "start_time": "2021-07-05T06:52:53.267821Z"
    }
   },
   "outputs": [],
   "source": [
    "documents = [TaggedDocument(doc, [i]) for i, doc in enumerate(common_texts)]\n",
    "model = Doc2Vec(documents, vector_size=10, window=5, min_count=1, workers=4, seed = 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-07-04T07:38:51.193311Z",
     "start_time": "2021-07-04T07:38:51.188919Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[['human', 'interface', 'computer'],\n",
       " ['survey', 'user', 'computer', 'system', 'response', 'time'],\n",
       " ['eps', 'user', 'interface', 'system'],\n",
       " ['system', 'human', 'system', 'eps'],\n",
       " ['user', 'response', 'time'],\n",
       " ['trees'],\n",
       " ['graph', 'trees'],\n",
       " ['graph', 'minors', 'trees'],\n",
       " ['graph', 'minors', 'survey']]"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "common_texts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-07-05T06:52:56.320055Z",
     "start_time": "2021-07-05T06:52:56.314615Z"
    }
   },
   "outputs": [],
   "source": [
    "# duplicated \"user\"\n",
    "duplicated_texts= [\n",
    "     ['human', 'interface', 'computer'],\n",
    "     ['survey', 'user', 'user','user', 'user','user', 'user','user', 'user','computer', 'system', 'response', 'time'],\n",
    "     ['eps', 'user', 'user','user', 'user','user', 'user','user', 'user','interface', 'system'],\n",
    "     ['system', 'human', 'system', 'eps'],\n",
    "     ['user', 'user','user', 'user','user', 'user','user', 'user','response', 'time'],\n",
    "     ['graph','trees'],\n",
    "     ['trees'],\n",
    "     ['graph', 'minors', 'trees'],\n",
    "     ['graph', 'minors', 'survey']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-07-05T06:53:01.476609Z",
     "start_time": "2021-07-05T06:53:01.458157Z"
    }
   },
   "outputs": [],
   "source": [
    "documents2 = [TaggedDocument(doc, [i]) for i, doc in enumerate(duplicated_texts)]\n",
    "model2 = Doc2Vec(documents2, vector_size=10, window=5, min_count=1, workers=4, seed = 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-07-05T06:53:02.108758Z",
     "start_time": "2021-07-05T06:53:02.100476Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([-0.01891889, -0.0091159 , -0.02980401,  0.02008216, -0.02158123,\n",
       "         0.0325665 ,  0.00983108,  0.03669152, -0.0015269 , -0.0073979 ],\n",
       "       dtype=float32),\n",
       " array([-0.01898973, -0.00911209, -0.02986462,  0.02012529, -0.02158619,\n",
       "         0.03259935,  0.00989106,  0.03665011, -0.00151132, -0.0074449 ],\n",
       "       dtype=float32))"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.wv['user'], model2.wv['user']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-07-05T06:53:06.996733Z",
     "start_time": "2021-07-05T06:53:06.992260Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[TaggedDocument(words=['human', 'interface', 'computer'], tags=[0]),\n",
       " TaggedDocument(words=['survey', 'user', 'user', 'user', 'user', 'user', 'user', 'user', 'user', 'computer', 'system', 'response', 'time'], tags=[1]),\n",
       " TaggedDocument(words=['eps', 'user', 'user', 'user', 'user', 'user', 'user', 'user', 'user', 'interface', 'system'], tags=[2]),\n",
       " TaggedDocument(words=['system', 'human', 'system', 'eps'], tags=[3]),\n",
       " TaggedDocument(words=['user', 'user', 'user', 'user', 'user', 'user', 'user', 'user', 'response', 'time'], tags=[4]),\n",
       " TaggedDocument(words=['graph', 'trees'], tags=[5]),\n",
       " TaggedDocument(words=['trees'], tags=[6]),\n",
       " TaggedDocument(words=['graph', 'minors', 'trees'], tags=[7]),\n",
       " TaggedDocument(words=['graph', 'minors', 'survey'], tags=[8])]"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "documents2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-07-05T06:53:08.662978Z",
     "start_time": "2021-07-05T06:53:08.480303Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(model.wv['user'], model2.wv['user'], 'r->', markersize = 20);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-07-05T06:53:13.489263Z",
     "start_time": "2021-07-05T06:53:13.360013Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(model.wv['survey'], model2.wv['survey'], 'r->', markersize = 20);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-07-05T06:53:18.306443Z",
     "start_time": "2021-07-05T06:53:18.152183Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAD+CAYAAADVsRn+AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nO3de3STZZ4H8O/7tk3KrbZJSktg5OIgLbB0gHLpkbRjydY9zooHZ125KHQZ8TB1quPMIisjrI5HpjMyU7XLRT2sIFNFHfDM6mFGqIfaMnjj0tCpYpWKSNNSkvRCr2mad/8ofacppW3SJG/y9vs5Z8745H2T/H4E+u3zXp4IkiRJICIiAiAqXQAREYUOhgIREckYCkREJGMoEBGRjKFAREQyhgIREckYCkREJIv05UknTpxAYWEhRFHE8uXLkZmZKW+7ePEiCgoK0NLSgtTUVGRnZ0MURRQUFODrr7+Gy+XCfffdh/T0dL81QURE/uF1KLS1tWH//v149tlnIYoiNm7ciNTUVMTExAAA9uzZg1WrViElJQVPP/00Tp48iYULF2L16tXQ6XSorq7G1q1bGQpERCHI68NHFosFycnJ0Ol0iI2NxezZs1FeXg4AaGpqQl1dHebOnQtRFGEymVBWVgYA0Ol0AIArV65g8uTJfmyBiIj8xeuZgs1mQ3x8vDzW6/Wor68HANjtdhgMBnmbTqfD6dOnAQClpaV47bXXIIoiNm3adMPXLyoqQlFREQAgLy8PTqfT2xLDQmRkJFwul9JlBAz7C19q7g1Qtj+NVgtJowFEEe41a9C1eTMwYYJ/30OjGdbzvQ4Fl8sFQRDksSAIEEWx322iKMrbTCYTTCYTzp07h+eeew6///3vMXr06Ote32w2w2w2y2ObzeZtiWHBYDCotjeA/YUzNfcGKNufEYBw7Rddce9eiPv2oeW++9D885/DnZDgn/cwGof1fK8PH8XFxcHhcMjj3rOD/rbp9XqP5yclJcFgMKCmpsbXmomIwp7gdELo6MCYAweQkJaGmCeegHj5stJleR8KKSkpsFgsaGxsRENDAyorKzFnzhwA3Qms1WpRUVEBt9uNkpISpKWlobGxEbW1tQCAmpoa2Gw2TPDzlImIKJQYsrKG9IM+1MJB8GXp7OLiYhw8eBAA8MADDwAAamtrsWzZMlRVVWHHjh1obW1FRkYGVqxYAbvdjm3btqGjowOjR4/G6tWrkZKSMqT3slqt3pYXFjhFD29q7k/NvQHB6884cWL3+QNBkA8RJc6bN+jz+j7H28NKwz185FMoBBNDITyxv/Cl5t6A4IZCj54f9EJHx5Cf72s4BP2cAhEReafnEJEvzwn2YSWGAhFRCAt2ODAUiIjCQLDCgaFARBRG5HDYvx9xOTl+f32fFsQjIiJl9NwR3XMC2t8YCkREYaBvGLjHjw/I+zAUiIhuwJCVBef8+X5dhsJbEgBERKD17rtxdfPmgIVBD55TICK6AU1FheJ3GgsA4HYj8rvvAh4IAEOBiGhASi5DIUVFQYqORsuaNajftSso78lQICIagmCGgxwGq1bh8kcfoWnbtqDMEgCGAhGRVwIZDkqGQQ+GAhGRD/wZDqEQBj149RER0TD0fGnOmAMHMObNN9F6zz0Y88YbQ3quFBUFREQE/DJTbzAUiIj8QA6HIQRCKIZBD4YCEVGQBOsGtOFgKBARBViwb0AbDp5oJiIKMAEAurqCdgPacHCmQEQUQH3PH4Q6hgIRUQCEw/mD/jAUiIj8KFzDoAdDgYjID8I9DHowFIiI+tPVNaTd1BIGPRgKRER9RP/5z9AN8lWXaguDHgwFIqIeHR0wTps24C5qDYMevE+BiAjAmJdf9giEht/9zmO7pNF0L1q3cqXii9YFEmcKRDSiCc3NmDBjhsdj1gsXgKgoxD7+uOpnBn0xFIhoxIp56imMfeUVeWx/9VV0ZGXJY+esWXCmpo6IMOjhUyicOHEChYWFEEURy5cvR2Zmprzt4sWLKCgoQEtLC1JTU5GdnQ1RFFFYWIjPPvsMHR0dSEtLw5o1a/zWBBGRN8QrV5D4gx/IY0kQUHPxIiB6HlG3HTkS7NIU5/U5hba2Nuzfvx/PPPMMnnnmGbzxxhtoamqSt+/ZswerVq3C//zP/+Dbb7/FyZMnAQAJCQn4wx/+gOeffx4WiwXnzp3zXxdEREMU+e//7hEIV/78Z9RcunRdIIxUXv8pWCwWJCcnQ6fTITY2FrNnz0Z5eTkAoKmpCXV1dZg7dy5EUYTJZEJZWRkAwGw2QxRFaLVaGI1GNDc3+7cTIqIBRHzzDYwTJ0L8858BAK5p02C9dAmdqakKVxZavD58ZLPZEB8fL4/1ej3q6+sBAHa7HQaDQd6m0+lw+vRpj+c3NDTgwoULePjhh/t9/aKiIhQVFQEA8vLyPF5PTSIjI1XbG8D+wpkae4tcsADi2bPy2GmxAElJUFeX/uF1KLhcLgiCII8FQYB4bdrVd5soivI2AHA6ncjPz8fatWsRHR3d7+ubzWaYzWZ5bLPZvC0xLBgMBtX2BrC/cKam3qLKyhD/ox/J4/bMTIh/+Ut3fyrpsS+j0Tis53sdCnFxcaioqJDHdrsd06dPl7c5HA6PbXq9HgDQ2dmJ7du3Iz09HamcrhFRIEkSEr//fYjt7fJDtadOwZ2YyNnBILw+p5CSkgKLxYLGxkY0NDSgsrISc+bMAdD9G4ZWq0VFRQXcbjdKSkqQlpYGl8uF/Px8LFiwAEuXLvV7E0REPbTHjsE4aZIcCC1r18JaXQ13YqLClYUHr2cKsbGxWLlyJZ588kkAwJo1a3D27FnU1tZi2bJlePjhh7Fjxw60trYiIyMDSUlJKCkpwZkzZ/Ddd9/h//7v/wAAOTk5SE5O9m83RDRydXXBePPNHg/VfPEFpJgYhQoKT4IkSZLSRQzEarUqXUJAqOm4bX/YX/gKx95GvfUW4h57TB43PfEEmn/2s373Dcf+vBH0cwpERCGjvR3GW27xeMh6/jxwgwtZaHC8W4OIwtLYggKPQKh//nlYq6sZCMPEmQIRhRWhsRETZs70eMz67bdAJH+c+QNnCkQUNm564gmPQLD/8Y/dswMGgt/wT5KIQp5YW4vE+fPlsXvMGNSeO8f1igKAf6JEFNJ099/vEQhXDh9GbWUlAyFAOFMgopAU+dVXGP/DH8pj5+zZsP31r0CvpXTI/xgKRBRaJAnjlyxB5IUL8kOXjx9H19SpytU0gnD+RUQhI+qzz2CcNEkOhLY774S1upqBEEScKRCR8iQJxkmTPB6qLSuDu9cy/RQcnCkQkaK0R454BELzQw91L2DHQFAEZwpEpIzOThinTPF4qObLLyGNHatMPQSAMwUiUsDoP/7RIxAan3oK1upqBkII4EyBiIJGaGvDhO9/3+Mx6zffABqNQhVRX5wpEFFQjNu+3SMQHDt3di9RwUAIKZwpEFFACfX1mDB7tsdj1osXgYgIhSqigXCmQEQBE/vYYx6BYHvrre7ZAQMhZHGmQER+F1FdjYSFC+VxV3w8Lp85wyUqwgBDgYj8Sv/jH0P78cfyuO7IEbhmzVKwIvIGQ4GI/CLyiy8w3myWxx0LF8J+6BBnB2GGoUBEwyNJSJg/HxGXL8sPXf7kE3T1WbaCwgNPNBORzzQnTsA4aZIcCK0//nH3AnYMhLDFmQIRec/thvF73/N4qLa8HG6dTqGCyF84UyAir0S/+65HIFx95JHuBewYCKrAmQIRDY3TCWOf7zWo+fprSKNGKVQQBQJnCkQ0qDF79ngEQsO2bd0L2DEQVCeoMwWXy4Xa2lpM4kkoorAgtLRgwq23ejxmvXABiIpSpiAKOJ9C4cSJEygsLIQoili+fDkyMzPlbRcvXkRBQQFaWlqQmpqK7OxsiKKIF198ERaLBVOmTMGWLVv81gARBca4Z5/FuJ075bFjzx60/8u/KFgRBYPXodDW1ob9+/fj2WefhSiK2LhxI1JTUxETEwMA2LNnD1atWoWUlBQ8/fTTOHnyJBYuXIglS5YgPT0d7777rt+bICI/qqu77soi63ffASKPNo8EXn/KFosFycnJ0Ol0iI2NxezZs1FeXg4AaGpqQl1dHebOnQtRFGEymVBWVgYAmDdvHqI45SQKaXEbNkDTKxBs77zTvYAdA2HE8HqmYLPZEN/ru1P1ej3q6+sBAHa7HQaDQd6m0+lw+vRpr16/qKgIRUVFAIC8vDyP11OTyMhI1fYGsL+wU1UFTXKyPJSmTkXnF18gRoVLVKjus/Mzr0PB5XJB6PUXRRAEiNd+i+i7TRRFedtQmc1mmHutn2Kz2bwtMSwYDAbV9gawv3BiuPNOaCwWeewsK4MtPh6w2xWsKnDU9Nn1x2g0Duv5Xs8J4+Li4HA45HHv2UF/2/R6/bAKJKLAiCovh3HiRDkQ2tPTuw8V9Zox0MjjdSikpKTAYrGgsbERDQ0NqKysxJw5cwB0J7BWq0VFRQXcbjdKSkqQlpbm96KJaBgkCYlJSYjvdSVR7WefwfHGGwoWRaHC68NHsbGxWLlyJZ588kkAwJo1a3D27FnU1tZi2bJlePjhh7Fjxw60trYiIyMDSUlJAIBNmzahoaEBLS0tyM3Nxb333ov09HT/dkNEA9KWlEC/cqU8blm9Go2/+52CFVGoESRJkpQuYiBWq1XpEgJC7cc12V+I6eqC8eabPR6q+fxzSDfddN2uYdebl9TeX9DPKRBReBl18KBHIDRt3Ni9REU/gUDEBfGI1KqjA8Zp0zwesp4/D0RHK1QQhQPOFIhUaMyuXR6BUP+HP3RfWcRAoEFwpkCkAENWFpzz56P55z+HOyHBb68rXL2KCdcu7uhh/fZbIJL/1GloOFMgUoCmogJjDhxAQloaYp54AmKv7zf2VcyWLR6BYH/tte7ZAQOBvMC/LUQKEZxOAMCYAwcw5s030XLffT7NHMTLl5E4b548lrRa1Hz9NdcrIp/wbw2RwgSnE0JHh08zB93atR6BcOW991BTVcVAIJ/xbw5RiPAmHCLOn4dx4kREX1s8sjMpCdZLl9A5d24wSyYVYigQhZgBw0GSEP/DHyKh12oAl0tLceWDDwAVrmhKwcdQIApRfcMh7sEHYZw0CVFffQUAaLvjDlirq9HV514EouHgiWaiENdzQnrUX/4iP1Z35Ahcs2YpVRKpGGcKRGFG0mgQf9ddfruUlag3hgJRmBnO1UpEg2EoEIUpORz270dcTo7S5ZBK8JwCUZiSNBpAFOWb3oj8gaFAFGb6hoF7/HilSyIVYSgQhQmGAQUDQ4EoxDEMKJgYCkRBNtQrhRgGpASGAlEQ6dasQfQHHwy4D8OAlMRQIAqCiPPnPdYr6g/DgEIBQ4EowOJvvx1RlZXy+PLx40hYskQeMwwolDAUiAIk6swZxP/rv8rjtjvvRP0rr8hjhgGFIoYCkb9JEiZMnQqhs1N+qLasDO74eHnsnDULztRUhgGFHIYCkR9pjx2D/v775XHzT36Cpl//+rr9bEeOBLMsoiFjKBD5Q1cXjDff7PFQzZdfQho7VqGCiHzDBfGIhmnU2297BELjk0/CWl3NQKCwxJkCka86OmDs861n1m++ATQahQoiGj6fQuHEiRMoLCyEKIpYvnw5MjMz5W0XL15EQUEBWlpakJqaiuzsbIiiiM8//xwvvfQSXC4Xli5dinvuucdvTRAFm7h9O4y/+pU8rn/xRbT9+McKVkTkH16HQltbG/bv349nn30Woihi48aNSE1NRUxMDABgz549WLVqFVJSUvD000/j5MmTWLBgAXbv3o1f/vKXSEhIwKZNmzBv3jxMmTLF3/0QBZRw9SomJCV5PGb97jtA5JFYUgevQ8FisSA5ORk6nQ4AMHv2bJSXl+O2225DU1MT6urqMHfuXACAyWRCWVkZ9Ho9brrpJkyePBkAsGjRIpSVlfUbCkVFRSgqKgIA5OXlwWAw+NpbSIuMjFRtb4A6+4t47DFE7NwpjzvffRdSVhbU1aU6P7ve1N7fcHkdCjabDfG9rrfW6/Wor68HANjtdo8/bJ1Oh9OnT8Nut1/3HKvV2u/rm81mmM1mj/dTI4PBoNreAGX6M2RlwTl/fve1/wkJfntdsa4Oidd+0QEAd3Q0XA0NsNntgAo/Q/7dDG9Go3FYz/d6zutyuSAIgjwWBAHitalz322iKEIUxQGfQ+QvmooKv39vcdy6dR6BcOXwYdSePw/0+vtMpCZe/2SOi4uDw+GQx71nB/1t0+v1iI2N9Xjc4XBAr9cPp26ifvnrS+0jvvkGxokTMer99wEAnUlJsFZXozMlxd8lE4UUr0MhJSUFFosFjY2NaGhoQGVlJebMmQOge1qm1WpRUVEBt9uNkpISpKWl4dZbb4XVaoXVakV7ezs++eQTLFq0yO/NEPUYTjjEm80eC9ZdLi3FlUGWuyZSC0GSJMnbJxUXF+PgwYMAgAceeAAAUFtbi2XLlqGqqgo7duxAa2srMjIysGLFCgBAWVkZXn31VbhcLtx9993Iysoa0nvd6NxDuFP7cU0l+jNOnHjDbZJGAwjCPxaf6+ecQ5TFgvg775THbVlZqH/11X5fT82fn5p7A9Tf33DPKfgUCsHEUAhPoRYKPfoNB0lC4vTpENva5P1qz5wZcKE6NX9+au4NUH9/QT/RTBTO+h5W0q1dC+OkSXIgNP/Hf8BaXc2VS2nE4jIXNCIJTicAIPraPTEAcLmkBF233KJUSUQhgTMFInQfVhr/z//st0tZicIVQ4FCkiErK6g/oP11KStRuGMoUEgKxI1oQyGHw/79iMvJCcp7EoUShgKFLCV+e5c0GkjR0WhZswb1u3YF9L2IQhFDgUJeMMJBDoOVK3H5o4/QtG0br0CiEYmhQGEjEOHAMCDyxEtSKez0XE465sABjHnzTflGNG9IGg0giv+4iY1BQASAoUBhrG84DAXDgGhgDAUKez3hMBCGAdHQMBRI1aSoKCAigmFANEQ80UwBFeyb0HqTAKCrC6133YXmRx9lIBANAUOBAkqpm9AAQAAguN0Y/ac/8UY0oiFiKFDAKbWEBG9EI/IeQ4GCJljhwHsPiHzHUKCgC1Q4MAyIho+hQIrxVzgwDIj8h5ekkuL63oQmdHQM6Xm894DI/xgKFDKGchMawDAgCiSGAoUNhgFR4DEUKORJ1/6/c+pUNBQUwDVrlqL1EKkZTzRTyBOu/S+qshI3bd2qdDlEqsaZAoW8voeNiChwGAoUsngOgSj4GAokM2RlwTl/fvcP4IQExepgGBApJ+ihYLVaYTAYoNFogv3WNAhNRQWivvrK49vMghkODAMi5XkdCg6HA/n5+bDZbJgxYwZycnI8fsB3dXVh9+7dOHv2LAwGAx599FGMHz8eJSUleOedd2C1WrFz507o9Xq/NkL+caOvuvQ6HDo6kPBP/zSkXRkGRKHD66uPXn/9dZhMJuzatQuiKOLo0aMe2z/88EN0dnZi9+7duP3227F3714AwIQJE7BlyxYYDAa/FE6BNZwlKEYdPAhNTAwiHI4B9+PyFEShx+uZwqlTp7B+/XoAQEZGBt577z386Ec/krd/+umnuPPOOyEIAkwmE/bt2wcAmD59+pBev6ioCEVFRQCAvLw81YZIZGRkWPTmMXN46y2416xB1+bNwIQJ1+/c1ARNfLw8dN9zD8RDh67bTdJogIgI+bU0iYnQBayDwAiXz88Xau4NUH9/w+VVKDQ3N0Oj0UCr1QIA9Ho9GhoaPPax2+2Iv/aDQavVQqvVorm5GWPHjh3Se5jNZpjNZnlss9m8KTFsGAyGkOvNOMC2nnAQ9+6FuG/fdYeVxrzyCm566il5f2d5OWw6HYy9QuGGh4lC7M9hKELx8/MXNfcGqL8/o3Ggf8mDGzQUqqqqkJ+fDwBITk6GKP7jiJMgCB5jAHC5XBAEYcB9KHz1PefQevfdGPPWW/L25nXr0PTMM92/iV37h8dzBkThY9BQmDZtGgoKCgAAbrcb2dnZcLlciIyMhN1uv+6EcWxsLBwOBxITE+F0OuF2uzF69OjAVE+KkcOhVyDU/fWvcPU5ueycNQvO1FSGAVGY8OpXeFEUMXPmTBw/fhwAUFxcjMWLF3vsM2/ePBQXFwMASkpKsGDBAv9USiFN0mgQf/fd152Qth05whPIRGHE6+M669atw9GjR7FhwwZoNBosWbIEbW1t+M1vfgO324077rgDTqcTP/3pT1FaWooVK1YAAA4dOoTc3Fw4HA5s3boVzz33nN+bIeX0vVopIjc3KN/DTET+JUiSJA2+m3KsVqvSJQREKJ7sMk6c6LfXkkQRzoULYT940G+vGUpC8fPzFzX3Bqi/v4CfaCbyRs9JZffatajfsEHpcojISwwF8ou+VxjpZs6EW8W/jRGpFUOBhoWXmxKpC0OBfMIwIFInhgIBAITW1iHtxzAgUjeGAiFu/XqMOnx4wH0YBkQjA0NhBBNtNiSmpAy4D8OAaGThokQjVPzSpR6B4HjpJY/tXNaaaGTiTGGEiTh/Hgnp6R6PWaur5f/mzIBoZGMojCATvvc9CG63PLa98w6cCxfKYy5eR0QMhREg6tQpxC9b5vFY79lBD9uRI8EqiYhCFENB5fquZ1T34Ydwff/7ClVDRKGOJ5pVKvr99z0CwTVtGqzV1QwEIhoQZwpqI0kwTprk8VDt6dPy12YSEQ2EMwUVGb13r0cgtJvNsFZXMxCIaMg4U1ADlwvGyZM9Hqr58ktIY8cqVBARhSvOFMLcuG3bPAKh+cEHYa2uZiAQkU84UwhTQmsrJkyf7vGY9cIFICpKmYKISBU4UwhDcQ895BEIjU891X3fAQOBiIaJM4Uw0t8CdtZLlwBBUKgiIlIbzhTCRLzZfN0CdtbqagYCEfkVZwohLqKqCgkmk8dj/S1RQUTkDwyFEDbh5pshdHXJY9uhQ3AuWqRgRUSkdgyFEDTUBeyIiPyNoRBirlvArrgYrj6XnhIRBQpPNHvBkJWFmCeegHj5st9f+7oF7KZM6V7AjoFAREHEmYIXNBUViPrqK4x5881/fDPZcNcV4gJ2RBRCvA4Fh8OB/Px82Gw2zJgxAzk5OdBoNPL2rq4u7N69G2fPnoXBYMCjjz6K8ePHo7y8HIWFhWhpaUFcXBweeeQRGAwGvzYTDILTCQAYc+DAsMNh9L59iN28WR63L10Kx2uv+a1WIiJveX346PXXX4fJZMKuXbsgiiKOHj3qsf3DDz9EZ2cndu/ejdtvvx179+4FALS0tGDz5s0oKCjAnDlz8Kc//ckvDShFcDohdHRgzIEDSEhL8+6wkssFjVbrEQg1584xEIhIcV6HwqlTp5CRkQEAyMjIQFlZmcf2Tz/9FJmZmRAEASaTCeXl5QCAxYsXIyYmBgAwdepUNDc3D7f2kOBtOFy3gN1PftK9gN24ccEol4hoQF4dPmpuboZGo4FWqwUA6PV6NDQ0eOxjt9sRHx8PANBqtdBqtWhubsbYXqt2lpSUYPHixf2+R1FREYqKigAAeXl5YXOIyeOw0ltvwb1mDbo2bwYmTOjeoaUFGp3O4znOq1eh0WgQHh16JzIyMmw+O1+ouT819waov7/hGjQUqqqqkJ+fDwBITk6GKP5jciEIgscYAFwuF4ReSy/03efw4cOQJAlLlizp9/3MZjPMZrM8ttlsQ2wl8IxD2KcnHMS9eyHu24eW++5DRHU1Rn3wgbxP43//N0Zt3hxSvfmbwWBgf2FKzb0B6u/PaBzKT6obGzQUpk2bhoKCAgCA2+1GdnY2XC4XIiMjYbfbodfrPfaPjY2Fw+FAYmIinE4n3G43Ro8eDQA4duwYzpw5g8cff3xYRYeDnnAY2+c8gfW77wBRxCgliiIiGoRX5xREUcTMmTNx/PhxAEBxcfF1h4HmzZuH4uJiAN2HiRYsWAAAKC0txd/+9jds3LgRUSNwiWcpIgKSVouYX/0qIPc5EBH5g9cnmtetW4ejR49iw4YN0Gg0WLJkCdra2vCb3/wGbrcbd9xxB5xOJ37605+itLQUK1asAADs2LED1dXV+OUvf4nc3Fz5qqSRQujq8jghHZGby3AgopAjSJIkKV3EQKxWq9IlyPouQTEckijCuXAh7AcP+u01Q4naj9uquT819waov7+An1Mg/5I0GkAU4V67FvUbNihdDhGRB4ZCkPSEQc8d0LqZM+FW8W8rRBSeGAoB1jcM3OPHK10SEdENMRQChGFAROGIoeBnDAMiCmcMhaHo7MT4H/5wwF0YBkSkBgyFQUQfPgzd+vU33M4wICI1YSjcgNDWhoTZsyG2twMA2tPTEV1SIm9nGBCRGjEU+jG6sBCxvdZnqjt6FK6ZM2GcOJFhQESqxlDoRWhowIRZs+Rx67/9GxpeeEEeO2fNgjM1lWFARKoVdqFgyMqCc/58/3w/ci9jX3wRMb/9rTy+/NFH6Lr5Zo99bEeO+O39iIhCkdcL4ilNU1Hh21dg3oBYUwPjxIlyIFz92c9gra6+LhCIiEaCsAsFYJjfj9xLzJYtSExNlce1FguuPvGEP0slIgorYRkKPXwNh4jz52GcOBFj//d/AXR/E5q1uhpufkUfEY1wYR0KPYYcDpKEuPXrkZCeLj9Uc+4cWh56KIjVEhGFLlWEQo+BwiHKYoFx0iSMOnwYAFD/4ouwVldDGjdOyZKJiEJK2F19NBQ934885sABjHnzTbjHjUPEtWWquwwGXP70U0CrVbJEIqKQpKqZQl89M4eeQGjPzMSVI0cYCEREN6DqUOhLe/y43y5lJSJSoxEVCv66lJWISK1GVCj0kMNh/37E5eQoXQ4RUchQ5YnmwfRd1I6IiLqNqFDgCqdERAMbEaHAMCAiGhpVhwLDgIjIO6oMBYYBEZFvVBUKDAMiouHxOhQcDgfy8/Nhs9kwY8YM5OTkQKPRyNu7urqwe/dunD17FgaDAY8++ijGjx+Pjz/+GG+//TZaW1uRlJSEhx9+GJGR/skkhgERkX94fZ/C66+/DpPJhF27dkEURRw9etRj+4cffojOzk7s3r0bt99+O/bu3QsAmDFjBrZv344dO3bAZrPBYrEMu3hJo4EUHY2WlStx+aOP0LRtGwOBiGgYvA6FU6dOISMjA+GXL6oAAAdxSURBVACQkZGBsrIyj+2ffvopMjMzIQgCTCYTysvLAQBxcXEQBAEtLS3o7OyE0Wj0uWiGARFRYHh1/Ka5uRkajQbaawvK6fV6NDQ0eOxjt9sRHx8PANBqtdBqtWhubkZUVBT+8z//EzabDffeey8mTJjQ73sUFRWhqKgIAJCXl3d9eEgShGv/Ofba/8LVcIIxHLC/8KXm3gD19zccg84UqqqqkJubi9zcXLz22msQxX88RRAEjzEAuFwuCIJw3T5arRYFBQV4+eWX8fe//x2lpaX9vp/ZbEZeXh7y8vLwX//1X772FfLU3BvA/sKZmnsD2N9gBp0pTJs2DQUFBQAAt9uN7OxsuFwuREZGwm63Q6/Xe+wfGxsLh8OBxMREOJ1OuN1ujB49Wt4+btw43HbbbaisrITJZBpW8URE5F9enVMQRREzZ87E8ePHAQDFxcVYvHixxz7z5s1DcXExAKCkpAQLFiwAAJw7dw5A90zi1KlTuOWWW4ZbOxER+VnEU0899ZQ3T7j11lvx+uuv4+2338bEiROxfPlytLe3Y/v27bjtttswbdo0fPzxx9i3bx/sdjvWr1+P6OhoHDp0CK+88go++OADJCUl4a677vI4zHQj06ZN87W3kKfm3gD2F87U3BvA/gYiSJIk+bEWIiIKYyPy+xSIiKh/DAUiIpIxFEgxVqsVTqdT6TJoiFwuFy5duqR0GRRgii6IF4rrKPmTr/2Vl5ejsLAQLS0tiIuLwyOPPAKDwaBgJ/3ztb+SkhK88847sFqt2Llz53WXNSvtxIkTKCwshCiKWL58OTIzM+VtFy9eREFBAVpaWpCamors7GyIoojPP/8cL730ElwuF5YuXYp77rlHwQ4G5kt/L774IiwWC6ZMmYItW7YoWP3gfOmvsLAQn332GTo6OpCWloY1a9Yo2MGN+dJbQUEBvv76a7hcLtx3331IT08f+E0kBRUUFEjvv/++JEmS9MILL0jvvfeex/YPPvhAys/Pl9xut3T06FHpt7/9rSRJkuRwOCS32y11dXVJTz75pHTy5Mmg1z4Uvvb30UcfSY2NjZIkSdLbb78t7dq1K7iFD5Gv/VVWVkp2u13KycmRbDZb0OseSGtrq7RhwwbJbrdL9fX10oMPPih/FpIkSVu3bpVOnz4tdXV1SVu3bpU++eQTye12S7m5udKFCxektrY26ZFHHpG++eYb5ZoYgC/9SZIknTp1Sjpz5oz061//WqnSh8TX/o4ePSp1dXVJ7e3t0i9+8Qvpiy++UKqFG/K1N7vdLkmSJF26dElat27doO+j6OGjUFhHKZB87W/x4sWIiYkBAEydOhXNzc3BLXyIfO1v+vTp0Ol0Qa93KCwWC5KTk6HT6RAbG4vZs2fLdTc1NaGurg5z586FKIowmUwoKytDVVUVbrrpJkyePBnR0dFYtGjRdX8WocKX/oDu+4+ioqKULH1IfO3PbDbLKy8YjcaQ/Dfna289/9auXLmCyZMnD/o+ioXCcNZR6ujoQG5uLh566CEsXLjwhusoKWk4/fVWUlJy3Q2CocBf/YUam80m1wx091VfXw+gu5/eh/F0Oh0aGho8+uz7nFDjS3/hZLj9NTQ04MKFC5g9e3ZwCvaCr72VlpZi/fr12LVrF+6///5B3yeoB+KrqqqQn58PAEhOTh72OkpXr15Ffn4+4uPjQ2LJDH/11+Pw4cOQJAlLliwJcOVD4+/+QtFANffdJooiRFEMqz596S+cDKc/p9OJ/Px8rF27FtHR0cEreoh87c1kMsFkMuHcuXN47rnn8Pvf/95j6aG+ghoKal9HyZ/9HTt2DGfOnMHjjz8e9D5uxN+fXyiKi4tDRUWFPLbb7Zg+fbq8zeFweGzT6/Vynz0cDkfInTzv4Ut/4cTX/jo7O7F9+3akp6cjNTU1uEUP0XA/u6SkJBgMBtTU1Ay4zJBivwaofR2l4fRXWlqKv/3tb9i4cWPIHscdTn+hLCUlBRaLBY2NjWhoaEBlZSXmzJkDADAYDNBqtaioqIDb7UZJSQnS0tJw6623wmq1wmq1or29HZ988gkWLVqkcCf986W/cOJLfy6XC/n5+ViwYAGWLl2qcAc35ktvjY2NqK2tBQDU1NTAZrMNerhd0WUu6urq8MILL8But+MHP/gBHnroIXR0dOD555/Hpk2b4HK5sHPnTnz55ZcYP348HnvsMcTGxuLll1/G6dOnodVqsWDBAqxevXpI6ygFm6/9rVixAnFxcfJltvPnz0d2drayzfTD1/4OHTqEY8eOwWazQafTYcqUKdi4caPS7ciKi4tx8OBBAMADDzwAAKitrcWyZctQVVWFHTt2oLW1FRkZGVixYgUAoKysDK+++ipcLhfuvvtuZGVlKVb/YHzpb9OmTWhoaJAvk7733nsHv7RRId72V1JSgl27dnkck8/JyUFycrIi9Q/E297sdju2bduGjo4OjB49GqtXr0ZKSsqA78G1j4iISBZeZ5GIiCigGApERCRjKBARkYyhQEREMoYCERHJGApERCRjKBARkYyhQEREsv8HJSlarIxXAd0AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(model.wv['computer'], model2.wv['computer'], 'r->', markersize = 20);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-07-05T06:53:27.008762Z",
     "start_time": "2021-07-05T06:53:27.002586Z"
    }
   },
   "outputs": [],
   "source": [
    "# 将词向量投影到二维空间\n",
    "rawWordVec = []\n",
    "word2ind = {}\n",
    "ind2word= {}\n",
    "for i, w in enumerate(model.wv.vocab):\n",
    "    rawWordVec.append(model[w])\n",
    "    word2ind[w] = i\n",
    "    ind2word[i] = w\n",
    "    \n",
    "rawWordVec = np.array(rawWordVec)\n",
    "\n",
    "rawWordVec2 = []\n",
    "word2ind2 = {}\n",
    "ind2word2= {}\n",
    "for i, w in enumerate(model2.wv.vocab):\n",
    "    rawWordVec2.append(model2[w])\n",
    "    word2ind2[w] = i\n",
    "    ind2word2[i] = w\n",
    "    \n",
    "rawWordVec2 = np.array(rawWordVec2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-07-05T06:53:27.922180Z",
     "start_time": "2021-07-05T06:53:27.472539Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXwAAAD6CAYAAACiefy7AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nO3df1hUdb4H8Pc5DDMopsAMPxy9lm7+QLmahiF3JQpZ7Vpra/dJxedGbD11De9WbllpP56bZHpvbbhLBD73+mgSiv2wrTV3V1kFNF3NHyDhFpapxQg4MyAxIMMwc/9ATgwKMsMwZ87M+/U8/vGd7+Hw+Tj69us5Z74IDofDASIi8nui3AUQEZF3MPCJiAIEA5+IKEAw8ImIAgQDn4goQDDwiYgChEruAm7EYDDIXYJLdDodjEaj3GV4XSD2zZ4Dh6f61o8aBYdaDQgCLIsXo/npp2GPjvZAhd2+h17f6xxX+EREXiRYrRDa2hBaVIToxEQMX7UKYl2dV743A5+ISAZyBD8Dn4hIRt4MfgY+EZEP8EbwM/CJiHzIYAY/A5+IyAcNRvAz8ImIfJgng5+BT0SkAJ4IfgY+EZGCSMFfUIDwzEyXvtbnP2lLREQ/cajVgChKn9R1BQOfiEgBega9PSrK5XMw8ImIfJgngr4LA5+IyAd5Mui7MPCJiLzBbu/XYYMR9F0Y+EREg0z9+efQLVrU5zGDGfRdGPhERIOlvR1Rd94J1YULvR7ijaDvwufwiYgGQchnn0F/yy1S2F/65BOneYdaDUdICCxpaag7fBhNr78+qGEPcIVPRORRQksLYqZMgWC1AgCupKTAvHUrIAgAvLui74mBT0TkIUMLChD2wgvSuH7fPtgmTpTG1ilTYI2P93rQd2HgExENkNDQgJFxcdLYkpaGy2++ec1xxj17vFnWNRj4REQDMCw7G8O7hXvdkSPoGD1axop6x8AnInKDePEiYuLjpfGPTz2FH597TsaKboyBT0TkohGrVyP03XelcW1lJewRETJW1D8MfCKiflJ98w2ikpOl8eWsLFgeeUTGilzjVuAfOnQIhYWFEEURCxcuREpKijR34cIF5OTkwGKxID4+HhkZGRDFnx73z8rKAgC8/PLLAyydiMhLHA6EP/oohvzlL9JLF6ur4QgNlbEo17n8wavW1lYUFBQgKysLWVlZ2L59O5qamqT5TZs2YenSpXj77bdx/vx5HDt2TJorKSmBSsX/VBCRcgSXl0MdEiKFfUNuLgw1NYoLe8CNwK+oqEBsbCwiIiIQFhaGuLg4VFZWAgCamppQX1+P6dOnQxRFJCUloby8XJrbt28f5s+f79kOiIgGg90O3X33IfLeewEAHTExMHz3HVp/9SuZC3Ofy8tto9GIyMhIaazVatHQ0AAAMJlM0Ol00lxERAROnDgBANiyZQvS0tJgv8GOccXFxSguLgYArF+/3ul8SqBSqRRXsycEYt/s2X8JxcUIvhr0AGD/85/RkZICpXfucuDbbDYIVz8iDACCIEjX6HvOiaIIURRRXl4OjUaD2NhYVFVV9Xn+1NRUpKamSmOj0ehqibLS6XSKq9kTArFv9uyHrFZEJyYiqLa2c3jbbTD+6U/QRUUppm+9Xt/rnMuXdMLDw2E2m6Vx91X99ea0Wi3279+Pr776CitXrkR+fj7OnDmDLVu2uPqtiYgGTcgnn0A/dqwU9pd27YLxs88A0X/2mHR5hT9t2jRs27YNly9fhsPhQHV1NR5//HEAnf/6azQaVFVVITY2FmVlZUhLS8OkSZOkr6+qqsLOnTuRkZHhsSaIiNwlWCyImTgRgsMBAGidNw8NmzZJm535E5cDPywsDGlpaXjppZcAAOnp6Th16hRqa2uxYMECLF++HLm5uWhpaUFycrJT2BMR+ZKhmzcj7GqWAUB9aSlst94qY0WDS3A4rv6z5qMMBoPcJbjE769x9iIQ+2bPyiWazYj553+Wxpb0dFxet67X45XUd1/X8PlQPBEFlJveeAM3bdggjWuPHoV91CgZK/IeBj4RBYSgmhpE33GHNG569lk0r1ghY0Xex8AnIr834rnnEFpYKI0vVlbCoYDNzjyNgU9EfktVXY2ou++Wxo3r1qElPV3GiuTFwCci2ejmzoX19ts7f+RfdLTnTuxwIOLhhxHyt791DoODUXv6NBxDh3rueyiQ/3yigIgUR11VhdCiIkQnJmL4qlUQ6+oGfM7gY8egHz1aCntzfj4unjsX8GEPMPCJSGaC1QqhrW3gwd/Rgci5cxF5//0AANs//RMM587hyi9/6eGKlYuBT0Q+YSDBr9m3D/oxYxB8da8uY1ER6v/+dyA4eDBLVhwGPhH5FJeCv60N0VOnQvvQQ53DmTNh+P57WJOSvFixcjDwicgn3Sj4h+zcCf24cQgymQAAl3bvhumPf/Srzc48jU/pEJFPE6xWAEBoURFCd+xAy8KFCC0qkuZb77sPDfn5frnZmacx8IlIEboHf5dLO3eiPSFBrpIUh//3ISJFcqjV0KWleexxzkDAwCciRfLY45wBhIFPRIomBX9BAcIzM+Uux6fxGj4RKZpDrQZEEZbFi9H89NNyl+PTGPhEpEg9g94eFSV3ST6PgU9EisKgdx8Dn4gUgUE/cAx8IpJHe3u/DmPQew4Dn4i8LjQ/HyOysvo8hkHveQx8IvIawWLByAkT+jyGQT94+Bw+EUl0c+cO2geYhr/yilPYm957z2neoVbDERICS1oa6g4fRtPrrzPsPYwrfCKSqKuqEHzmDEJ37PhphT3AHz0oGo2ImTZNGjs0Glw8e/anMVf0XsMVPhE58eSWBeGPPuoU9pc++8wp7K1TpnBF70Vc4RPRdfXcltiyeDGwZk2/fopU0HffIXr2bGncHhuLS8XF1xxn3LPHcwXTDXGFT0R96r7iD5406YYr/sjUVKewrzt48LphT97HwCeifhGsVghXrvR6qSe4ogL6UaMQ/I9/AABa77kHhpoadIwdK1fJ1AMv6RCRS653qSe0qEh6HQBqy8thj4yUq0TqBVf4ROSWrks9w7ZulcLekpYGQ00Nw95HcYVPRB7hCA7G0J074QgO9sjjnOR5bgX+oUOHUFhYCFEUsXDhQqSkpEhzFy5cQE5ODiwWC+Lj45GRkQFRFFFYWIgvvvgCbW1tSExMRHp6useaIKJr6ebORZDBgCupqfhx1apBD2Dh6t443S/1MPh9i8uXdFpbW1FQUICsrCxkZWVh+/btaGpqkuY3bdqEpUuX4u2338b58+dx7NgxAEB0dDTeeustbNiwARUVFfjqq6881wURXUNdVYWghgYM/eADRM+ciRFPP+2VHwHIn0Dlu1wO/IqKCsTGxiIiIgJhYWGIi4tDZWUlAKCpqQn19fWYPn06RFFEUlISysvLAQCpqakQRREajQZ6vR7Nzc2e7YSIrksAIHR0eC34pS0S0tPRkJc3aN+HXOfyJR2j0YjIbjdktFotGhoaAAAmkwk6nU6ai4iIwIkTJ5y+vrGxEefOncPy5cuve/7i4mIUX31md/369U7nUwKVSqW4mj0hEPtWWs8CAFwN/qE7d8KeloaO114DRo70yPkdajUQFAR7ejo6Vq+GOiYGER45s/yU9l73xuXAt9lsEARBGguCAFEUrzsniqI0BwBWqxXZ2dl4+OGHERISct3zp6amIjU1VRobjUZXS5SVTqdTXM2eEIh9+3rP+l5e7wp+8b33IG7fjpYHHpCu8ff2NX3pdS8cH/69cZWvv9fd6fW9v4suB354eDiqqqqksclkwvjx46U5s9nsNKfVagEA7e3tePPNN3HnnXciPj7e1W9LRB7Wc8Xf8sADLn09Nz1THpev4U+bNg0VFRW4fPkyGhsbUV1djalTpwLo/FdQo9GgqqoKdrsdZWVlSExMhM1mQ3Z2NmbOnIk5c+Z4vAkicl/3a/z9wW2MlcvlFX5YWBjS0tLw0ksvAQDS09Nx6tQp1NbWYsGCBVi+fDlyc3PR0tKC5ORkTJo0CWVlZTh58iS+//57fPrppwCAzMxMxMbGerYbInKbcIP5rmv0lkWLuKJXKMHhcDjkLqIvBoNB7hJcoqRrfZ4UiH37es/6UaM8ch6HKALBwbAsWQL1mjUwqgLv85q+/l5319c1fG6tQER9s9uB9nYILS2Ab68P6QYY+ETUJwGAYLdj6AcfQPXv/y53OTQADHwi6pMDgCMoCC2LFsFWWCh3OTQAgXcxjoj6xQEAQUFo+bd/63xOPyqq88NHCrmWTddi4BORk+sFPfkHBj4RAWDQBwIGPpEfCn/iiX4fy6APHAx8Ij8imkyIufrJ9xth0AceBj6Rn4j8xS8QfPr0DY9j0AcuBj6RwgV99x2iZ892es1QU3PNJ20Z9MTAJ1KwkWPGQOjokMbGjz+G9Y47nI5h0FMXBj6RAgUfP47IBQucXjPU1DiNrVOmdP5M21/8gkFPABj4RIrT81JNfWkpbLfees1xxj17vFUSKQS3ViBSiJC//MUp7Nt/9jMYamquG/ZE18MVPpGvczigHz3a6aXakyd5iYZcxhU+kQ8bunmzU9i33nMPDDU1DHtyC1f4RL7IZoP+5pudXrpYXQ1HaKhMBZE/4AqfyMcMX7PGKeybly2DoaaGYU8DxhU+kY8QLBaMnDDB6TXD+fNAAP5IQRocXOET+YDwRx5xCvvGtWs7n6tn2JMH8U8TUT/p5s6F9fbb0fz007BHR3vknGJ9PWKmT3d6zfDDD4AgeOT8RN1xhU/UT+qqKoQWFSE6MRHDV60CLl4c0Pki77zTKezNmzZ1ruoZ9jRIGPhELhCsVghtbQgtKkLwpEkYvmoVxLo6l86h+uYb6EeNQvC330qvGWpqcOWeezxdLpETBj6RGwSrFcKVK04r/v4Ev37UKEQlJ0vjS59+es0eOESDhYFPNADdV/x9Bb/6yBGnbREcKhUMNTVov/12b5ZLAY43bYk8QLBaAQChRUUI3bEDlsWLpZu7PTc7qzt4EB1jx8pRJgU4Bj6RBzkF//btENrbpbn2KVNwiTtYkox4SYdoEAhWq1PYWxYtgqmgQMaKiBj4RF4x9I9/dOnmLtFgYOATeUF/b+4SDSYGPpEXScFfUIDwzEy5y6EA49ZN20OHDqGwsBCiKGLhwoVISUmR5i5cuICcnBxYLBbEx8cjIyMDoiji9OnT2LhxI2w2G+bMmYMHHnjAY00QKYVDrQZEUXqKh8ibXA781tZWFBQUYO3atRBFEStXrkR8fDyGDx8OANi0aROWLl2KadOm4dVXX8WxY8cwc+ZM5Ofn45lnnkF0dDSef/55zJgxA7fccoun+yHyST2Dnj/AhOTgcuBXVFQgNjYWERERAIC4uDhUVlbi5z//OZqamlBfX4/pV/cHSUpKQnl5ObRaLUaMGIGbr+7xnZCQgPLy8usGfnFxMYqLiwEA69evh06nc7c3WahUKsXV7AmB2veNONRqICgI9vR0dKxeDXVMDCLkLmoAAvV99pe+XQ58o9GIyMhIaazVatHQ0AAAMJlMTr8pEREROHHiBEwm0zVfYzAYrnv+1NRUpKamOn0/JdHpdIqr2RN8pe/B2NGyi96FY3td0fvA79FA+Mr77G1K6luv7/1Pqss3bW02G4Ruu/kJggBRFK87J4oiRFHs82uIPKnnjpaeeBJGffDgNZ+W7Y1DrYYjJASWtDTUHT6Mptdf5+Ub8hkur/DDw8NRVVUljU0mE8aPHy/Nmc1mpzmtVouwsDCn181mM7Ra7UDqJupVX9scuMqVoOc1evJ1Li+zp02bhoqKCly+fBmNjY2orq7G1KlTAXT+t0ej0aCqqgp2ux1lZWVITEzEhAkTYDAYYDAYcOXKFRw5cgQJCQkeb4aou4E8+z6kqKhfYc8VPSmJ4HA4HK5+UUlJCT766CMAwEMPPQQAqK2txYIFC3D27Fnk5uaipaUFycnJWLJkCQCgvLwcmzdvhs1mw/3334+5c+f263v1dq3fVynpWp8n+UrffYW0Q60GBOGGK/6e56j/619hi4tz3u3y6s1Yy6JFAbWi95X32duU1Hdf1/DdCnxvYuArg6/03d9V+fWC/6bXX8dNublOx3bfq14/apTTpRv1mjUwBtjPnPWV99nblNR3X4EfWH9aiXCda/yLFmFYj43Nao8dg33kSKfXrFOmwBofL63odTqd4p+6ocDCwCdFGIzHLbuCv3vY24cNQ+3XX1/3eCO3NiaF47ORpAiD8bhlT47gYAjt7dzYjPwWA58UY7B3nBTa27mjJfk1Bj4pzqAHP3e0JD/Fa/ikWNf7gJUncEdL8lcMfFK87sE/EPy0LPk7Bj75ja7gdxWDngIFA59kM5g7W/YHg54CDQOfZKOuqkLwmTMD3uDMVQx6ClR8Sodk5c0f7u24+qt97Fhc+vRTbnRGAYeBTz7BG8EvXP0VXF2NEa+84tFzEykBA598ymAGv7SVcXo6GvLyPHJOIiXhNXzyST2fsR8IXrMn6sTAJ5/m7qOWAIOeqCcGPvkdBj3R9THwyW8w6In6xsAnxXMAQFAQWu6/Hz+uXs2gJ+oFn9IhxRMAwG6H6vvvGfZEfeAKnxTNERzc+cPEubMl0Q0x8EmRegY9V/ZEN8bAJ0Vh0BO5j4FPisAncIgGjoFPsgj69tt+HcegJ/IcBj55l8OB8Mcfx5Ddu/s+jEFP5HEMfPKa4IoKRM6f3+cxDHqiwcPAp8Fnt0P3q19Bffw4AKBDp0Pd0aPQjxsnHcKgJxp8DHwaVOqyMujS0qSxqaAAbSkp0phBT+Q9DHwaHO3tiJo9G6offgAAWOPiYNy9GwgKkg6xTpkCa3w8g57ISxj45HEhf/oTIpYtk8aXPv0U7bfffs1xxj17vFkWUcBj4JPHCC0tiImNhWCzAQCuzJkD87vvAoIgc2VEBLgR+GazGdnZ2TAajZg4cSIyMzOhVqul+Y6ODuTn5+PUqVPQ6XR46qmnEBUVhcrKShQWFsJisSA8PBxPPvkkdDqdR5sh+YgbN2Lkk09K4/r9+2GbMEHGioioJ5d3y9y2bRuSkpKQl5cHURSxd+9ep/nS0lK0t7cjPz8fd999N7Zs2QIAsFgsWL16NXJycjB16lR8+OGHHmmA5CWYzdCPGgXV1bC3LF0KQ00Nw57IB7m8wj9+/Dgee+wxAEBycjJ27dqFe++9V5o/evQo5s+fD0EQkJSUhHfffRcAMGvWLOmYsWPH4ty5c9c9f3FxMYqLiwEA69evV9z/AlQqleJqdpf42mtQZWVJY+uZMwgeMwaB0X1gvdddArFnwH/6dinwm5uboVarodFoAABarRaNjY1Ox5hMJkRGRgIANBoNNBoNmpubMWzYMOmYsrIyp38AuktNTUVqaqo0NhqNrpQoO51Op7iaXSXW1CDmjjuk8Y8rVkCzfn1n337ee3eB8F73FIg9A8rqW6/X9zp3w8A/e/YssrOzAQCxsbEQxZ+uAgmC4DQGAJvNBqHbTbqex+zevRsOhwOzZ8/ufwfkM0asWoXQrVulcW1lJewREdDIWBMR9c8NA3/cuHHIyckBANjtdmRkZMBms0GlUsFkMkGr1TodHxYWBrPZjJiYGFitVtjtdgwdOhQAsH//fpw8eRLPPffcILRCg0l15gyi7rpLGje+9hpafv1r+QoiIpe5dNNWFEVMnjwZBw8eBACUlJRcc2lmxowZKCkpAdB56WbmzJkAgAMHDuDzzz/HypUrERwc7IHSySscDoT/+tdS2DsEARerqxn2RArk8lM6jzzyCPbu3Ytly5ZBrVZj9uzZaG1txbp162C32zFv3jxYrVY88cQTOHDgAJYsWQIAyM3NRU1NDZ555hn85je/kZ7eId8VfOIE9KNHY8jVD0iZ33kHF3/4AY7QUJkrIyJ3CA6HwyF3EX0xGAxyl+ASJd3c6VVHB3T33Qf1qVOdw5gY1B0+DHT7vEVPftG3i9hz4FBS333dtHV5hU/+TVNSAv2YMVLYm7ZvR93x432GPREpA7dWoE5WK6JnzUJQXV3ncPp0GD/9FBC5JiDyF/zbTAj55BPox46Vwv7Srl0w7trFsCfyM1zhBzChuRkjJ06Uxq3/+q9o+N//5WZnRH6KS7gAFbppk1PY15eWouH//o9hT+THuMIPMKLJhJipU6WxJT0dl9etk7EiIvIWBn4Auem//xs3/eEP0rj2iy9g7+MRLiLyLwz8ABD0ww+ITkiQxk3PPovmFStkrIiI5MDA93Mjnn0Wodu3S+OLX34JR3i4jBURkVwY+H5K9fXXiEpJkcaN69ahJT1dxoqISG4MfH/jcCDioYcQsn9/51CjQe2XX8JxdcdSIgpcfCzTjwR/8QX0o0dLYW/euBEXz55l2BMRAK7w/UNHByLnzUPwP/4BALCNGYP6sjKA21ATUTdc4Suc5m9/g37MGCnsjTt2oP7wYYY9EV2DK3ylamtDdHw8gszmzuEdd8D00Ufc/4aIesV0UKAhH34I/bhxUthf+vOfYfr4Y4Y9EfWJK3wFEX78ESMnTZLGrb/8JRry8rj/DRH1C5eEChG6caNT2NcdOICG/HyGPRH1G1f4Pk68dAkxt90mjZsffRRNa9bIWBERKRUD34fdtHYtbnrnHWlce+IE7NHRMlZERErGwPdBQRcuIDoxURo3vfACmn/zGxkrIiJ/wMD3MWFPPomhH30kjS+ePg3HiBEyVkRE/oKB7yNUVVWImjtXGje++SZa0tJkrIiI/A0DX24OB7RLlkBz8CAAwB4aitqKCmDIEJkLIyJ/w8cyZaQ+cgT60aOlsDdv2oTa6mqGPRENCq7w5WCzITI1FcFnzgAA2n/2M1zatw9Q8e0gosHDFb6Xafbsgf7mm6WwN374IS6VlTHsiWjQMWW8pbUVMTNmQGxqAgC0/cu/wPT++/ykLBF5DVf4XjBkxw7ob71VCvv6v/4Vpg8+YNgTkVdxhT+IhMuXMXLyZGncsnAhGt9+W8aKiCiQMfAHybB33sHwtWulcd3nn6PjllvkK4iIAp7LgW82m5GdnQ2j0YiJEyciMzMTarVamu/o6EB+fj5OnToFnU6Hp556ClFRUdJ8aWkpcnNzkZeXB61W65kufMnFi9B3C/bm//gPNL3yinz1EBFd5fI1/G3btiEpKQl5eXkQRRF79+51mi8tLUV7ezvy8/Nx9913Y8uWLdJcU1MT9u/fD51ON+DCfdHwV1+FulvY1548ybAnIp/h8gr/+PHjeOyxxwAAycnJ2LVrF+69915p/ujRo5g/fz4EQUBSUhLeffddaW7r1q1YtGgRcnNzez1/cXExiouLAQDr169Xxj8O33wD9ZQp0tD2+uuwP/MMImQsydtUKpUy3isPYs+Bw1/6dinwm5uboVarodFoAABarRaNjY1Ox5hMJkRGRgIANBoNNBoNmpub8e2330KlUmFyt5uY15OamorU1FRpbDQaXSnR68IyMzH0k0+ksbW+Hsb2dsDH6/Y0nU7n8++Vp7HnwKGkvvV6fa9zNwz8s2fPIjs7GwAQGxsLsdvPTRUEwWkMADabDUK3xw27jnn//fexatUql4v3Vaovv0TUvHnSuOGtt9C6eDF0I0YEXNgTkTLcMPDHjRuHnJwcAIDdbkdGRgZsNhtUKhVMJtM1N17DwsJgNpsRExMDq9UKu92OyspKNDY24tVXXwXQeeN37dq1WLt2LYYobd8Yux3aBx+E5u9/7xyGhaH2+HEgJETmwoiI+ubSJR1RFDF58mQcPHgQd911F0pKSjBr1iynY2bMmIGSkhJMnjwZZWVlmDlzJhISEpCQkCAds3z5crz44ouKC3v1oUPQPfigNDZt3oy2blsaExH5Mpef0nnkkUewd+9eLFu2DGq1GrNnz0ZrayvWrVsHu92OefPmwWq14oknnsCBAwewZMmSwajbu9rbEfXzn0th3z5pEgznzzPsiUhRBIfD4ZC7iL4YDAZZv3/I7t2IuPpUEgAYP/4Y1jvu6PV4Jd3c8aRA7Js9Bw4l9T2gm7aBSmhtRXRcHMQrVwAAV5KTYS4s5P43RKRY3DztOoa+9x5G3nqrFPb1xcUwb9vGsCciReMKvxuhsREju32AquXBB9G4YYOMFREReY7iVvi6uXMxfNUqiHV1Hj3vsN//3ins6w4fZtgTkV9RXOCrq6oQWlSE6MREjwS/ePEi9KNGYfj//A8A4Mf//E8YamrQMWaMJ8olIvIZigt8ABCsVghtbQMO/uEvvYSY+HhpXFtRgR/96NPARETdKTLwu7gb/EHffAP9qFEYtnkzAODyf/0XDDU1sPvB5khERL1RdOB36XfwOxwIf+wxRCcnSy9d/PprWLo9Z09E5K/8IvC79BX8wRUV0I8ejSG7dwMAGv7wBxhqauAYNkzOkomIvMYvH8sUrFYAQGhREUJ37ID9ppsQdPVTch06HeqOHgWubvFMRBQo/GqF31PXir8r7K+kpODSnj0MeyIKSH4d+D1pDh702OOcRERKE1CB76nHOYmIlCigAr+LFPwFBQjPzJS7HCIir/DLm7Y34lCrAVGEZfFiND/9tNzlEBF5RUAFfs+gt0dFyV0SEZHXBETgM+iJiPw88Bn0REQ/8cvAZ9ATEV3LrwKfQU9E1Du/CHwGPRHRjSk68Bn0RET9p8jAZ9ATEblOcDgcDrmLICKiwReQWysMphdeeEHuEmQRiH2z58DhL30z8ImIAgQDn4goQDDwPSw1NVXuEmQRiH2z58DhL33zpi0RUYDgCp+IKEAw8ImIAgQDn4goQCjyk7ZyM5vNyM7OhtFoxMSJE5GZmQm1Wi3Nd3R0ID8/H6dOnYJOp8NTTz2FqG6fBi4tLUVubi7y8vKg1WrlaMFl7vZcWVmJwsJCWCwWhIeH48knn4ROp5Oxk/45dOgQCgsLIYoiFi5ciJSUFGnuwoULyMnJgcViQXx8PDIyMiCKIk6fPo2NGzfCZrNhzpw5eOCBB2TswD3u9F1YWIgvvvgCbW1tSExMRHp6uowduM6dnrtkZWUBAF5++WWv1+0OrvDdsG3bNiQlJSEvLw+iKGLv3r1O86WlpWhvb0d+fj7uvvtubNmyRZpramrC/v37FRF63bnbs8ViwerVq5GTk4OpU6fiww8/lKF615UXd1EAAAR8SURBVLS2tqKgoABZWVnIysrC9u3b0dTUJM1v2rQJS5cuxdtvv43z58/j2LFjcDgcyM/Px29/+1v87ne/Q2lpKc6dOydfE25wp28AiI6OxltvvYUNGzagoqICX331lVwtuMzdngGgpKQEKpWy1swMfDccP34cycnJAIDk5GSUl5c7zR89ehQpKSkQBAFJSUmorKyU5rZu3YpFixZBEASv1jxQ7vY8a9YsDB8+HAAwduxYNDc3e7dwN1RUVCA2NhYREREICwtDXFyc1E9TUxPq6+sxffp0iKKIpKQklJeX4+zZsxgxYgRuvvlmhISEICEh4ZrfI1/nTt9A5yOLoihCo9FAr9cr4j3u4m7PTU1N2LdvH+bPny9n+S5j4LuoubkZarUaGo0GAKDVatHY2Oh0jMlkQmRkJABAo9FAo9GgubkZFRUVUKlUmDx5stfrHoiB9NxdWVkZZs2a5Z2iB8BoNEq9AJ39NjQ0AOjss/v/ziIiItDY2OjUf8+vUQp3+u6usbER586dQ1xcnHcK9gB3e96yZQvS0tIUt8JXVrUyOXv2LLKzswEAsbGxTtfwBEFwGgOAzWZzWsF3HfP+++9j1apV3il6gDzVc5fdu3fD4XBg9uzZg1z5wPXVS885URQhiuIN+1cCd/ruYrVakZ2djYcffhghISHeK3qA3Om5vLwcGo0GsbGxqKqq8nrNA8HA74dx48YhJycHAGC325GRkQGbzQaVSgWTyXTNjdewsDCYzWbExMTAarXCbrejsrISjY2NePXVVwF03gRdu3Yt1q5diyFDhni9pxvxRM9Dhw4FAOzfvx8nT57Ec8895/U+3BEeHu70F9lkMmH8+PHSnNlsdprTarVS/13MZrNibsh3cadvAGhvb8ebb76JO++8E/Hx8d4teoDc6Xn//v24cOECVq5ciStXruDy5cvYsmULMjIyvF2+y5S1BPEBoihi8uTJOHjwIIDOGzc9L1PMmDEDJSUlADovY8ycORMJCQnIzc3FG2+8gTfeeAMRERF48cUXfTLse3K3ZwA4cOAAPv/8c6xcuRLBwcFerdtd06ZNQ0VFBS5fvozGxkZUV1dj6tSpAACdTgeNRoOqqirY7XaUlZUhMTEREyZMgMFggMFgwJUrV3DkyBEkJCTI3Ilr3OnbZrMhOzsbM2fOxJw5c2TuwHXu9LxixQpkZ2fjjTfewLJlyzB+/HhFhD3ArRXcUl9fj9///vcwmUy47bbb8Pjjj6OtrQ0bNmzA888/D5vNhnfeeQdff/01oqKisGLFCoSFhTmdY/ny5VizZo1iVoHu9rxkyRKEh4dL1zpvv/12RfzlKCkpwUcffQQAeOihhwAAtbW1WLBgAc6ePYvc3Fy0tLQgOTkZS5YsAQCUl5dj8+bNsNlsuP/++zF37lzZ6neXq32XlZUhLy/P6Vp3ZmYmYmNjZanfHe68112qqqqwc+dOxTyWycAnIgoQvKRDRBQgGPhERAGCgU9EFCAY+EREAYKBT0QUIBj4REQBgoFPRBQgGPhERAHi/wHnlo01HmFiAgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(rawWordVec[0], rawWordVec2[0], 'r->', markersize = 20);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-07-04T07:31:52.232213Z",
     "start_time": "2021-07-04T07:31:51.916754Z"
    }
   },
   "outputs": [
    {
     "ename": "ValueError",
     "evalue": "too many values to unpack (expected 2)",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-11-044b28ee5395>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      5\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      6\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mi\u001b[0m \u001b[0;32min\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m'survey'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'computer'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'eps'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'interface'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'response'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 7\u001b[0;31m     \u001b[0mx1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my1\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mrawWordVec\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mword2ind\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'user'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      8\u001b[0m     \u001b[0mx2\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my2\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mrawWordVec\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mword2ind\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      9\u001b[0m     \u001b[0max\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mplot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mx1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx2\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0my1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my2\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'r--'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mValueError\u001b[0m: too many values to unpack (expected 2)"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXwAAAFnCAYAAACl9vKSAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAZoElEQVR4nO3dX0jUe/7H8dd8f+rIqU6jM0VEaLhtZphWdLa80GFtCA4bnQqCCAqD3Vhit+jEoSDI/nCoWA4msx3r4tCGVLsXQiznZmk8mYhQnGI8UcuJcD1zxHXNGTUs7ew0399FnNnsj/r135if5+Oqme989P397p5n42em0WXbti0AwIxnpXoAAMDUIPgAYAiCDwCGIPgAYAiCDwCGIPgAYAiCDwCGmJTgJxIJRSKRyfjSAIAxShvLoubmZl2+fFmWZWnLli0qLy9PHqutrVVTU5MyMjIUDAaT9x87dkyPHz+WZb38O6a6ujr5ZwDA5HMc/IGBAdXW1urzzz+XZVn67LPPtGbNGn344YeSpKKiIvn9fv3pT396Y21lZaXmz58//qkBAI45ford0tKigoICZWdny+PxqLCwUPfu3UseLy4uVmZm5oQOCQAYP8fP8Lu7uzVv3rzkba/Xq56enhHXpaen6/jx45o1a5Y2btyosrKytz4uFAopFApJkk6fPu10PADAOzgOfjwel8vlSt52uVyj2os/cuSIJKm9vV0nTpzQkiVLtHDhwjceFwgEFAgEnI4FABiB4y2drKwsxWKx5O1oNCqfzzfq9YsWLVJ+fr7a29udfmsAwDg4Dn5xcbFaWlrU19en3t5ePXz4UEVFRSOu6+zslCQ9fvxYjx490uLFix0PCwAYO9dYPg+/oaFBdXV1kqSdO3dKehn0TZs26cyZM4pEIorFYvL5fFq/fr02b96sgwcPanBwUG63W9u2bVNJScnEngkAYFhjCj4A4P3Dv3wCAEOM6V/aAsCEO3pUev0jWXJypBMnUjPPDETwAUwPkYj0+ps52tpSMcmMxZYOABiC4AOAIQg+ABiCPXwA00NOzpt79jk5KRllpuJ9+ABgCLZ0AMAQBB8ADEHwAcAQBB8ADEHwAcAQM+9tmXweBwC81cwLPp/HAQBvxZYOABiC4AOAIQg+ABhi5u3h83kcAPBWfJYOABiCLR0AMATBBwBDEHwAMATBBwBDEHwAMATBBwBDEHwAMATBBwBDEHwAMATBBwBDEHwAMATBBwBDEHwAMATBBwBDEHwAMATBBwBDzLzfeAVgch09KkUiQ+/LyZFOnEjNPBg1gg/AmUhEWrx46H2v/1pRTEts6QCAIQg+ABiC4AOAISZlDz+RSKi9vV05OTmT8eUBpFJOzpt79vy3/l5w2bZtO13U3Nysy5cvy7IsbdmyReXl5cljtbW1ampqUkZGhoLBYPL+Bw8e6MKFC4rH41q/fr22bt06MWcAABgVx1s6AwMDqq2t1cmTJ3Xy5EldvXpVT548SR4vKirSkSNHhqyxbVvnz5/Xp59+qi+++EI3b95UG6/qA8CUchz8lpYWFRQUKDs7Wx6PR4WFhbp3717yeHFxsTIzM4esaW1t1dy5c5Wbm6vMzEytXbtW4XB4/NMDAEbN8R5+d3e35s2bl7zt9XrV09Mz7JpoNPrGmo6Ojrc+NhQKKRQKSZJOnz7tdDwAwDs4Dn48HpfL5Uredrlcsqzhf1BwsiYQCCgQCDgdCwAwAsdbOllZWYrFYsnb0WhUPp9v2DUej2fImlgsJq/X6/RbAwDGwXHwi4uL1dLSor6+PvX29urhw4cqKioads3SpUvV0dGhjo4ODQ4O6tatW1q7du2YhwYAODemt2U2NDSorq5OkrRz505JUmdnpzZt2qQzZ84oEokoFovJ5/Np/fr12rx5s8LhsC5evKh4PK5PPvlEGzZsmNgzAQAMa0zBBwC8f/hoBQAwBMEHAEMQfAAwBMEHAEMQfAAwBMEHAEMQfAAwBMEHAEMQfAAwBMEHAEMQfAAwBMEHAEMQfAAwhOPfeAVgkhw9KkUiQ+/LyZFOnEjNPJhxCD4wXUQi0uLFQ+9ra0vFJJih2NIBAEMQfAAwBMEHAEOwhw9MFzk5b+7Z5+SkZBTMTPxOWwAwBM/wkRq8BRGYcgQfqcFbEIEpx4u2AGAIgg8AhmBLB8DE4vWZaYvgIzV4C+LMxesz0xbBR2rwbA+YcuzhA4AhCD4AGIItHQATi9dnxm6SX/Am+AAmFq/PjN0kv+DNlg4AGILgA4AhCD4AGII9fACYLib5BW8+Dx8ADMGWDgAYguADgCEIPgAYguADgCEIPgAYguADgCHG9D785uZmXb58WZZlacuWLSovL08ei0QiCgaDevr0qdasWaOKigpZlqVjx47p8ePHsqyXf8dUV1cn/wwAmHyOgz8wMKDa2lp9/vnnsixLn332mdasWaMPP/xQkvTVV19px44dKi4u1vHjx/Xtt9/qV7/6lSSpsrJS8+fPn9gzAACMiuOn2C0tLSooKFB2drY8Ho8KCwt17949SdKTJ0/U1dWlVatWybIslZaWKhwOO/r6oVBIhw8f1uHDh52OBgAYhuNn+N3d3Zo3b17yttfrVU9PjyQpGo3K5/Mlj2VnZ+vu3buSpPT0dB0/flyzZs3Sxo0bVVZW9tavHwgEFAgEnI4FABiB4+DH43G5XK7kbZfLldyLf/2YZVnJY0eOHJEktbe368SJE1qyZIkWLlw4ruEBAKPneEsnKytLsVgsefvVZ/VvO+b1eoesX7RokfLz89Xe3j7WmQEAY+A4+MXFxWppaVFfX596e3v18OFDFRUVSZJ8Pp/cbrfu37+vRCKhxsZGlZSUSJI6OzslSY8fP9ajR4+0+PXf6gIAmFRj+rTMhoYG1dXVSZJ27twp6WXQN23apNbWVp07d07Pnj2T3+/X9u3bJUkHDx7U4OCg3G63tm3blvyLAAAwNfh4ZAAwBP/yCQAMQfABwBAEHwAMQfABwBAEHwAMQfABwBAEHwAMQfABwBAEHwAMQfABwBAEHwAMQfABwBAEHwAMQfABwBAEHwAMQfABwBAEHwAMQfABwBAEHwAMQfABwBAEHwAMQfABwBAEHwAMQfABwBAEHwAMQfABwBAEHwAMQfABwBAEHwAMQfABwBAEHwAMQfABwBAEHwAMQfABwBAEHwAMQfABwBAEHwAMQfABwBAEHwAMQfABwBAEHwAMkTaWRc3Nzbp8+bIsy9KWLVtUXl6ePBaJRBQMBvX06VOtWbNGFRUVsixLDx480IULFxSPx7V+/Xpt3bp1wk4CADAyx8/wBwYGVFtbq5MnT+rkyZO6evWqnjx5kjz+1VdfaceOHfrzn/+sH374Qd9++61s29b58+f16aef6osvvtDNmzfV1tY2kecBABiB4+C3tLSooKBA2dnZ8ng8Kiws1L179yRJT548UVdXl1atWiXLslRaWqpwOKzW1lbNnTtXubm5yszM1Nq1axUOhyf8ZAAA7+Z4S6e7u1vz5s1L3vZ6verp6ZEkRaNR+Xy+5LHs7GzdvXtX0Wj0jTUdHR1v/fqhUEihUEiSdPr0aafjAQDewXHw4/G4XC5X8rbL5ZJlWW89ZlmWLMsads3rAoGAAoGA07EAACNwvKWTlZWlWCyWvP3qs/q3HfN6vfJ4PEPuj8Vi8nq945kbAOCQ4+AXFxerpaVFfX196u3t1cOHD1VUVCRJ8vl8crvdun//vhKJhBobG1VSUqKlS5eqo6NDHR0dGhwc1K1bt7R27doJPxkAwLu5bNu2nS5qaGhQXV2dJGnnzp2SpM7OTm3atEmtra06d+6cnj17Jr/fr+3bt0uSwuGwLl68qHg8rk8++UQbNmyYwNMAAIxkTMEHALx/+Je2AGAIgg8AhiD4AGAIgg8AhiD4AGAIgg8AhiD4AGAIgg8AhiD4AGAIgg8AhiD4AGAIgg8AhiD4AGAIgg8AhiD4AGAIgg8AhiD4AGAIgg8AhiD4AGAIgg8AhiD4AGAIgg8AhiD4AGAIgg8AhiD4AGAIgg8AhiD4AGAIgg8AhiD4AGAIgg8AhiD4AGAIgg8AhiD4AGAIgg8AhiD4AGAIgg8AhiD4AGAIgg8AhiD4AGAIgg8AhiD4AGAIgg8AhkhzuiAWi6mqqkrd3d3Kz8/X3r17lZGRkTz+4sULnT9/Xt999518Pp/279+v+fPnq6GhQZcuXdLs2bMlSTt27FBJScnEnQkAYFiOn+FfuXJFpaWlqqmpkWVZun79+pDjN2/e1H//+1+dP39ev/71r/WXv/wleezjjz9WMBhUMBgk9gAwxRwH/86dO/L7/ZIkv9+vcDg85Pjt27dVXl4ul8ul0tJS3bt3b2ImBQCMi6Mtnf7+fmVkZMjtdkuSvF6vent7hzwmGo1q3rx5kiS32y23263+/n65XC7V19erqalJhYWF2rVrlzIzM9/4HqFQSKFQSJJ0+vTpMZ0UAOBNIwa/tbVVVVVVkqSCggJZ1v9+KHC5XENuS1I8HpfL5XrjMX6/X36/X4ODg6qpqdG1a9e0ffv2N75fIBBQIBAY8wkBAN5uxODn5eUpGAxKkhKJhCoqKhSPx5WWlqZoNCqv1zvk8R6PR7FYTAsWLNBPP/2kRCKhDz74IHk8MzNTZWVlqq+vn+BTAQAMx9EevmVZWr58uZqamiRJDQ0NWrdu3ZDHrF69Wg0NDZKkxsZGffTRR5Kkzs5O2batFy9eqLm5WUuWLJmA8QEAo+Wybdt2sqCrq0vV1dWKRqNauXKl9uzZo+fPn+vs2bM6dOiQ4vG4vvzyS33//feaP3++Dhw4II/Ho7q6OoVCIaWlpWnFihXavXu30tPTJ+u8AACvcRx8AMD7iX9pCwCGIPgAYAiCDwCGIPgAYAiCDwCGIPgAYAiCDwCGIPgAYAiCDwCGIPgAYAiCDwCGIPgAYAiCDwCGIPgAYAiCDwCGIPgAYAiCDwCGIPgAYAiCDwCGIPgAYAiCDwCGIPgAYAiCDwCGIPgAYAiCDwCGIPgAYAiCDwCGIPgAYAiCDwCGIPgAYAiCDwCGIPgAYAiCDwCGIPgAYAiCDwCGIPgAYAiCDwCGIPgAYAiCDwCGIPgAYAiCDwCGmLTgd3d3q7+/f7K+PADAoTSnC2KxmKqqqtTd3a38/Hzt3btXGRkZyePhcFhXr15VJBJRZWWlli1bJkkaGBhQdXW12tratGjRIu3fv19z5syZuDMBAAzL8TP8K1euqLS0VDU1NbIsS9evXx9yPCsrSwcOHFB+fv6Q+//+979r0aJFOn/+vH7xi1+orq5ufJMDABxxHPw7d+7I7/dLkvx+v8Lh8JDjubm5WrBgwRvrbt++rUAg8M51AIDJ5WhLp7+/XxkZGXK73ZIkr9er3t7eUa2NRqPy+XzJdT09PW99XCgUUigUkiSdPn3ayXgAgGGMGPzW1lZVVVVJkgoKCmRZ//uhwOVyDbk9nHg8LpfLNeK6QCCQ/EkAADBxRgx+Xl6egsGgJCmRSKiiokLxeFxpaWmKRqPyer2j+kYej0c9PT3y+XyKxWLJZ/sAgKnhaA/fsiwtX75cTU1NkqSGhgatW7duVGtXr16tb775RpJ048aNUa8DAEwMl23btpMFXV1dqq6uVjQa1cqVK7Vnzx49f/5cZ8+e1aFDh9TY2Ki6ujrFYjHNmTNHWVlZOnXqlPr7+1VdXa0ff/xReXl52rdvnzIzMyfrvAAAr3EcfADA+4mPVgAAQxB8ADAEwQcAQxB8ADAEwQcAQxB8ADAEwQcAQxB8ADAEwQcAQxB8ADAEwQcAQxB8ADAEwQcAQxB8ADAEwQcAQxB8ADAEwQcAQ4z4S8yRQkePSpHI0PtycqQTJ1IzD4D3GsGfziIRafHiofe1taViEgAzAFs6AGAIgg8AhmBLxzS8LgAYi+BPZzk5b+7Z5+SM72vyugBgLII/nfGsG8AEYg8fAAwx85/hs2cNAJJMCD571kNNxusCAN4LMz/4GIqfbABjsYcPAIYg+ABgiJm/pcOeNQBIkly2bdupHgIAMPlm/jN8wDS8FRnvQPCBmYa3IuMdeNEWAAxB8AHAEAQfAAzBHj4w0/BWZLwDb8sEAEOwpQMAhiD4AGCISQt+d3e3+vv7J+vLAwAccvyibSwWU1VVlbq7u5Wfn6+9e/cqIyMjeTwcDuvq1auKRCKqrKzUsmXLJEkNDQ26dOmSZs+eLUnasWOHSkpKJug0AAAjcfwM/8qVKyotLVVNTY0sy9L169eHHM/KytKBAweUn5//xtqPP/5YwWBQwWCQ2APAFHP8DP/OnTv63e9+J0ny+/36+uuv9Zvf/CZ5PDc3d+KmA8aDz5QBhnAU/P7+fmVkZMjtdkuSvF6vent7R7XW5XKpvr5eTU1NKiws1K5du5SZmfnG40KhkEKhkCTp9OnTTsYDhuIzZYAhRgx+a2urqqqqJEkFBQWyrP/tArlcriG3h+P3++X3+zU4OKiamhpdu3ZN27dvf+NxgUBAgUBgtPMDAEZpxODn5eUpGAxKkhKJhCoqKhSPx5WWlqZoNCqv1+voG2ZmZqqsrEz19fVjmxgAMCaOXrS1LEvLly9XU1OTpJfvvFm3bt2o1nZ2dsq2bb148ULNzc1asmSJ82kBAGPm+KMVurq6VF1drWg0qpUrV2rPnj16/vy5zp49q0OHDqmxsVF1dXWKxWKaM2eOsrKydOrUKdXV1SkUCiktLU0rVqzQ7t27lZ6ePlnnBfCiLfAaPksHAAzBRysAgCEIPgAYguADgCEIPgAYguADgCEIPgAYguADgCEIPgAYguADgCEIPgAYguADgCEI/msOHz6c6hHGhLmn1vs6t/T+zs7c40fwAcAQBB8ADPF/x44dO5bqIaabvLy8VI8wJsw9td7XuaX3d3bmHh8+Dx8ADMGWDgAYguAPo7u7W/39/akew7H3dW7gXRKJhCKv/7rK98B0mzst1QOkQiwWU1VVlbq7u5Wfn6+9e/cqIyMjeTwcDuvq1auKRCKqrKzUsmXLJL38pe2XLl3S7NmzJUk7duxQSUnJtJ97YGBA1dXVamtr06JFi7R//37NmTNn2sz94sULnT9/Xt999518Pp/279+v+fPnp/R6Nzc36/Lly7IsS1u2bFF5eXnyWCQSUTAY1NOnT7VmzRpVVFTIsiw9ePBAFy5cUDwe1/r167V169YpmXW8cx87dkyPHz+WZb18/lddXZ3883SYu7a2Vk1NTcrIyFAwGEzeP92v97vmTun1tg0UDAbtf/zjH7Zt23Z1dbX99ddfDzne1tZm//vf/7YrKyvtf/7zn8n7b9y4Yf/tb3+b0llfNda5//rXv9q1tbW2bdv2lStX7IsXL07ZzLY98tz19fV2VVWVnUgk7OvXr9tnzpyxbTt11/vZs2f273//ezsajdo9PT32b3/7W7uvry95/OjRo/bdu3ftFy9e2EePHrVv3bplJxIJ+49//KPd1tZmDwwM2Pv27bP/9a9/Tfu5bdu2Kysr7f/85z9TOuurRpo7HA7bP/zwg/2HP/whed/7cL3fNrdtp/Z6G7mlc+fOHfn9fkmS3+9XOBwecjw3N1cLFixIxWjDGuvct2/fViAQeOe6yTbS3Ldv31Z5eblcLpdKS0t17969KZ3vdS0tLSooKFB2drY8Ho8KCwuTMz158kRdXV1atWqVLMtSaWmpwuGwWltbNXfuXOXm5iozM1Nr166d8us8lrmng+HmlqTi4mJlZmYOWTPdr/e75k4147Z0+vv7lZGRIbfbLUnyer3q7e0d1VqXy6X6+no1NTWpsLBQu3btmrL/QcczdzQalc/nS67r6emZtDlfN5q5o9Go5s2bJ0lyu91yu93q7+9P2fXu7u5OzvPzzD9fs1evpSRlZ2fr7t27Q87h5zUdHR2TPuurxjK3JKWnp+v48eOaNWuWNm7cqLKysmkz97tM9+s9nFRebyOC39raqqqqKklSQUHBkP0yl8s16v0zv98vv9+vwcFB1dTU6Nq1a9q+ffukzCxN3NzxeFwul8vxurFyOver8736mKm+3iPN87ZjlmXJsqxh10yVscwtSUeOHJEktbe368SJE1qyZIkWLlw4LeaeyDUTbawzpPJ6GxH8vLy85IsmiURCFRUVisfjSktLUzQaldfrdfT1MjMzVVZWpvr6+skYN2mi5vZ4POrp6ZHP51MsFhvyTG8yOJ3b4/EoFotpwYIF+umnn5RIJPTBBx8kj0/V9f5ZVlaW7t+/n7wdjUb1y1/+MnksFosNOeb1epPn8LNYLOb4/1fjNZa5X7Vo0SLl5+ervb19SoM/3NzvMt2v92ik4nobt4dvWZaWL1+upqYmSS/febNu3bpRre3s7JRt23rx4oWam5u1ZMmSyRx1iPHMvXr1an3zzTeSpBs3box63UQYzdyrV69WQ0ODJKmxsVEfffSRpNRd7+LiYrW0tKivr0+9vb16+PChioqKJEk+n09ut1v3799XIpFQY2OjSkpKtHTpUnV0dKijo0ODg4O6deuW1q5dOyXzjmdu6eV1lqTHjx/r0aNHWrx48bSZ+12m+/UeTiqvt5H/0rarq0vV1dWKRqNauXKl9uzZo+fPn+vs2bM6dOiQGhsbVVdXp1gspjlz5igrK0unTp1SXV2dQqGQ0tLStGLFCu3evVvp6enTfu7+/n5VV1frxx9/VF5envbt2zelLyaNNHc8HteXX36p77//XvPnz9eBAwfk8XhSer0bGhpUV1cnSdq5c6ekl/+hbtq0Sa2trTp37pyePXsmv9+f3GYKh8O6ePGi4vG4PvnkE23YsGFKZh3v3AcPHtTg4KDcbre2bds2pW81Hs3cZ86cUSQSSf50un79em3evHnaX+93zZ3K621k8AHARMZt6QCAqQg+ABiC4AOAIQg+ABiC4AOAIQg+ABiC4AOAIQg+ABiC4AOAIf4f5Th0HIfAYggAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize = (6, 6))\n",
    "ax = fig.gca()\n",
    "ax.set_facecolor('white')\n",
    "ax.plot(rawWordVec[:, 0], rawWordVec[:, 1], 's', markersize = 5, alpha = 0.5, color = 'red')\n",
    "\n",
    "for i in ['survey', 'computer', 'eps', 'interface', 'response']:\n",
    "    x1, y1 = rawWordVec[word2ind['user']]\n",
    "    x2, y2 = rawWordVec[word2ind[i]]\n",
    "    ax.plot([x1, x2], [y1, y2], 'r--')\n",
    "\n",
    "# 设置中文字体，否则无法在图形上显示中文\n",
    "#zhfont1 = matplotlib.font_manager.FontProperties(fname='/Library/Fonts/华文仿宋.ttf', size=26)\n",
    "for k in ind2word:\n",
    "    x,y = rawWordVec[k]\n",
    "    if ind2word[k] in ['survey', 'computer', 'eps', 'interface', 'response', 'user']:\n",
    "        plt.text(x, y, ind2word[k], alpha = 1, color = 'purple', fontsize = 16, rotation=0)\n",
    "    else:\n",
    "        plt.text(x, y, ind2word[k], alpha = 1, color = 'black', fontsize = 16, rotation=0)\n",
    "\n",
    "plt.axis('off');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-07-04T07:31:52.233796Z",
     "start_time": "2021-07-04T07:31:50.608Z"
    }
   },
   "outputs": [],
   "source": [
    "fig = plt.figure(figsize = (6, 6))\n",
    "ax = fig.gca()\n",
    "ax.set_facecolor('white')\n",
    "ax.plot(rawWordVec2[:, 0], rawWordVec2[:, 1], 's', markersize = 5, alpha = 0.5, color = 'red')\n",
    "\n",
    "for i in ['survey', 'computer', 'eps', 'interface', 'response']:\n",
    "    x1, y1 = rawWordVec2[word2ind['user']]\n",
    "    x2, y2 = rawWordVec2[word2ind[i]]\n",
    "    ax.plot([x1, x2], [y1, y2], 'r--')\n",
    "\n",
    "# 设置中文字体，否则无法在图形上显示中文\n",
    "#zhfont1 = matplotlib.font_manager.FontProperties(fname='/Library/Fonts/华文仿宋.ttf', size=26)\n",
    "for k in ind2word2:\n",
    "    x,y = rawWordVec2[k]\n",
    "    if ind2word2[k] in ['survey', 'computer', 'eps', 'interface', 'response', 'user']:\n",
    "        plt.text(x, y, ind2word2[k], alpha = 1, color = 'purple', fontsize = 16, rotation=0)\n",
    "    else:\n",
    "        plt.text(x, y, ind2word2[k], alpha = 1, color = 'black', fontsize = 16, rotation=0)\n",
    "\n",
    "plt.axis('off');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-07-05T06:53:40.199839Z",
     "start_time": "2021-07-05T06:53:38.881120Z"
    }
   },
   "outputs": [],
   "source": [
    "tsne = TSNE(n_components=2, n_iter=10000000)\n",
    "X_tsne = tsne.fit_transform(rawWordVec)\n",
    "\n",
    "tsne2 = TSNE(n_components=2, n_iter=10000000)\n",
    "X_tsne2 = tsne2.fit_transform(rawWordVec2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-07-05T06:53:40.963930Z",
     "start_time": "2021-07-05T06:53:40.878595Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize = (6, 6))\n",
    "ax = fig.gca()\n",
    "ax.set_facecolor('white')\n",
    "ax.plot(X_tsne[:, 0], X_tsne[:, 1], 's', markersize = 5, alpha = 0.5, color = 'red')\n",
    "\n",
    "for i in ['survey', 'computer', 'eps', 'interface', 'response']:\n",
    "    x1, y1 = X_tsne[word2ind['user']]\n",
    "    x2, y2 = X_tsne[word2ind[i]]\n",
    "    ax.plot([x1, x2], [y1, y2], 'r--')\n",
    "\n",
    "# 设置中文字体，否则无法在图形上显示中文\n",
    "#zhfont1 = matplotlib.font_manager.FontProperties(fname='/Library/Fonts/华文仿宋.ttf', size=26)\n",
    "for k in ind2word:\n",
    "    x,y = X_tsne[k]\n",
    "    if ind2word[k] in ['survey', 'computer', 'eps', 'interface', 'response', 'user']:\n",
    "        plt.text(x, y, ind2word[k], alpha = 1, color = 'purple', fontsize = 16, rotation=0)\n",
    "    else:\n",
    "        plt.text(x, y, ind2word[k], alpha = 1, color = 'black', fontsize = 16, rotation=0)\n",
    "\n",
    "plt.axis('off');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-07-05T06:53:42.181238Z",
     "start_time": "2021-07-05T06:53:42.103729Z"
    },
    "code_folding": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 2\n",
    "fig = plt.figure(figsize = (6, 6))\n",
    "ax = fig.gca()\n",
    "ax.set_facecolor('white')\n",
    "ax.plot(X_tsne2[:, 0], X_tsne2[:, 1], 's', markersize = 5, alpha = 0.5, color = 'red')\n",
    "\n",
    "for i in ['survey', 'computer', 'eps', 'interface', 'response']:\n",
    "    x1, y1 = X_tsne2[word2ind2['user']]\n",
    "    x2, y2 = X_tsne2[word2ind2[i]]\n",
    "    ax.plot([x1, x2], [y1, y2], 'r--')\n",
    "\n",
    "# 设置中文字体，否则无法在图形上显示中文\n",
    "#zhfont1 = matplotlib.font_manager.FontProperties(fname='/Library/Fonts/华文仿宋.ttf', size=26)\n",
    "for k in ind2word2:\n",
    "    x,y = X_tsne2[k]\n",
    "    if ind2word2[k] in ['survey', 'computer', 'eps', 'interface', 'response', 'user']:\n",
    "        plt.text(x, y, ind2word2[k], alpha = 1, color = 'purple', fontsize = 16, rotation=0)\n",
    "    else:\n",
    "        plt.text(x, y, ind2word2[k], alpha = 1, color = 'black', fontsize = 16, rotation=0)\n",
    "\n",
    "plt.axis('off');\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![](img/word2vec/w2v.png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": false,
   "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": 4
}