archived-files/3_topic_modeling.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "## ISRC Python Workshop 3: Topic Modeling (Apr. 06th 2017)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "Text data is a great source for data analysis: sentiment analysis, topic retrieval, and so force. In this section, we will be covering a common precedure to conduct some basic analyses on text data retrieved from Scopus API. If you are interested in how to setup and get your own Twitter scraper work, please refer to the <a href=\"https://nbviewer.jupyter.org/github/zhiyzuo/uiowa-isrc-python/blob/master/2_web_scraping.ipynb\">previous IPython Notebook</a> for more details on web data retrieval."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true,
    "solution2": "hidden",
    "solution2_first": true
   },
   "source": [
    "### 0. Retrieving Text Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true,
    "solution2": "hidden"
   },
   "outputs": [],
   "source": [
    "## suppress warnings\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true,
    "solution2": "hidden"
   },
   "outputs": [],
   "source": [
    "with open(\"data/scopus_apikey.csv\", \"r\") as f:\n",
    "    key = f.read().strip()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true,
    "solution2": "hidden"
   },
   "outputs": [],
   "source": [
    "from pyscopus.scopus import Scopus\n",
    "## Supply your own key here.\n",
    "scopus = Scopus(key)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true,
    "solution2": "hidden"
   },
   "outputs": [],
   "source": [
    "## retrieve scopus ids and titles\n",
    "results = scopus.search_venue(count = 5000, year_range=(2014,2016), \\\n",
    "                              sort_by='relevancy', show=False, \\\n",
    "                              venue_title='Journal of Machine Learning Research')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true,
    "scrolled": true,
    "solution2": "hidden"
   },
   "outputs": [],
   "source": [
    "abstracts = []\n",
    "for item in results:\n",
    "    abstracts.append(scopus.retrieve_abstract(scopus_id=item))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "### Text Mining"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "Specifically, three popular clustering algorithms have been used to mine the underlying topical structures:\n",
    "\n",
    "(i) <a href=\"https://www.datascience.com/blog/introduction-to-k-means-clustering-algorithm-learn-data-science-tutorials\" target=\"_blank\">K-means</a>\n",
    "\n",
    "(ii) Latent Semantic Analysis (LSA) [1]\n",
    "\n",
    "(iii) <span style=\"color:red\">Latent Dirichlet Allocation (LDA)</span> [2]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "We will not cover WHAT these algorithms are, but it is good to understand the basics of the assumptions and the mechanisms."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "###### In this workshop, we will use implementations of LDA mentioned above from package <a href=\"http://scikit-learn.org/stable/\" target=\"_blank\">scikit-learn</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "### 1. Preprocessing Text Data (Data Cleaning)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "Before we do anything with the fancy algorithms, we must keep in mind that the most important step is to clean the data. Real-world data has many noises in it. We may get results that do not make any sense without proper preprocessing."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "## Let's look at the data first\n",
    "with open('data/corpus.csv', 'r') as f:\n",
    "    corpus = [text.replace('\\r', '').replace('\"','') \\\n",
    "              for text in f.read().split('\\n')\\\n",
    "              if len(text) > 0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Many real-world applications require varying costs for different types of mis-classification errors. Such a cost-sensitive classification setup can be very different from the regular classification one, especially in the multiclass case. Thus, traditional meta-algorithms for regular multiclass classification, such as the popular one-versus-one approach, may not always work well under the cost-sensitive classification setup. In this paper, we extend the one-versus-one approach to the field of cost-sensitive classification. The extension is derived using a rigorous mathematical tool called the cost-transformation technique, and takes the original one-versus-one as a special case. Experimental results demonstrate that the proposed approach can achieve better performance in many cost-sensitive classification scenarios when compared with the original one-versus-one as well as existing cost-sensitive classification algorithms. Reduction from cost-sensitive multiclass classification to one-versus-one binary classification\n",
      "--------------\n",
      "Variational inference has become a widely used method to approximate posteriors in complex latent variables models. However, deriving a variational inference algorithm generally requires significant model-specific analysis. These efforts can hinder and deter us from quickly developing and exploring a variety of models for a problem at hand. In this paper, we present a black box variational inference algorithm, one that can be quickly applied to many models with little additional derivation. Our method is based on a stochastic optimization of the variational objective where the noisy gradient is computed from Monte Carlo samples from the variational distribution. We develop a number of methods to reduce the variance of the gradient, always maintaining the criterion that we want to avoid difficult model-based derivations. We evaluate our method against the corresponding black box sampling based methods. We find that our method reaches better predictive likelihoods much faster than sampling methods. Finally, we demonstrate that Black Box Variational Inference lets us easily explore a wide space of models by quickly constructing and evaluating several models of longitudinal healthcare data. Black box variational inference\n",
      "--------------\n",
      "BayesPy is an open-source Python software package for performing variational Bayesian inference. It is based on the variational message passing framework and supports conjugate exponential family models. By removing the tedious task of implementing the variational Bayesian update equations, the user can construct models faster and in a less error-prone way. Simple syntax, flexible model construction and efficient inference make BayesPy suitable for both average and expert Bayesian users. It also supports some advanced methods such as stochastic and collapsed variational inference. BayesPy: Variational Bayesian inference in Python\n",
      "--------------\n"
     ]
    }
   ],
   "source": [
    "## Show 3 abstracts\n",
    "for doc in corpus[:3]:\n",
    "    print(doc)\n",
    "    print('--------------')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "source": [
    "#### Bag of words\n",
    "We do not really care the order of the words. We consider each document a <a href=\"https://en.wikipedia.org/wiki/Bag-of-words_model\" target=\"_blank\">bag of words</a>."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[u'2008',\n",
       " u'2011',\n",
       " u'2013',\n",
       " u'97',\n",
       " u'able',\n",
       " u'access',\n",
       " u'account',\n",
       " u'accuracy',\n",
       " u'accurate',\n",
       " u'achieve',\n",
       " u'achieves',\n",
       " u'active',\n",
       " u'actually',\n",
       " u'acyclic',\n",
       " u'adapt']"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "## scikit-learn provides convenient functions to preprocess data\n",
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "## remove stop words, which are some common words like I, you, and we.\n",
    "counter = CountVectorizer(max_df=0.95, min_df=2, \\\n",
    "                                stop_words='english')\n",
    "corpus_counter = counter.fit_transform(corpus)\n",
    "counter.get_feature_names()[:15]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "917"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(counter.get_feature_names())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "However, this seems not reasonable: (i) we probably do not want numbers; (ii) words in different forms are redundant (`achieve` and `achieves` are basically the same thing)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "## first remove numbers\n",
    "from string import digits\n",
    "corpus_nodigits = [doc.translate(None, digits) for doc in corpus]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "Let's try the `CountVectorizer` again."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "## stemming: get the `stem` of the words\n",
    "from nltk.stem.porter import PorterStemmer\n",
    "stemmer = PorterStemmer()\n",
    "corpus_for_stem = [doc.split() for doc in corpus_nodigits]\n",
    "corpus_stemmed_nodigits = [' '.join([stemmer.stem(word) for word in doc])\\\n",
    "                            for doc in corpus_for_stem]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(80, 789)\n",
      "(789,)\n"
     ]
    }
   ],
   "source": [
    "counter = CountVectorizer(max_df=0.95, min_df=2, \\\n",
    "                                stop_words='english')\n",
    "## corpus_counter is a doc-term sparse matrix\n",
    "corpus_counter = counter.fit_transform(corpus_stemmed_nodigits)\n",
    "print(corpus_counter.shape)\n",
    "words = np.asarray(counter.get_feature_names())\n",
    "print(words.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[u'abl',\n",
       " u'access',\n",
       " u'account',\n",
       " u'accur',\n",
       " u'accuraci',\n",
       " u'accuracy',\n",
       " u'achiev',\n",
       " u'activ',\n",
       " u'actual',\n",
       " u'acycl',\n",
       " u'ad',\n",
       " u'adapt',\n",
       " u'addit',\n",
       " u'addition',\n",
       " u'additionally']"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "counter.get_feature_names()[:15]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "### 2. Some summary statistics (Know your data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[array([u'cost', u'classif', u'sensit', u'approach', u'multiclass'], \n",
      "      dtype='<U14'), array([u'variat', u'model', u'infer', u'method', u'box'], \n",
      "      dtype='<U14'), array([u'variat', u'bayesian', u'model', u'construct', u'support'], \n",
      "      dtype='<U14')]\n",
      "[array([[8, 8, 6, 3, 3]]), array([[7, 6, 5, 5, 4]]), array([[5, 4, 2, 2, 2]])]\n"
     ]
    }
   ],
   "source": [
    "## Top 5 words of each document and their frequencies\n",
    "## NOTE: toarray is not recommended. This is just for illustration purpose.\n",
    "top_word_indices = np.argsort(-corpus_counter.toarray())[:, :5]\n",
    "top_words = [words[top_word_indices[index,:]] \\\n",
    "             for index in range(top_word_indices.shape[0])]\n",
    "top_word_frequency = [corpus_counter[index, top_word_indices[index]].toarray() \\\n",
    "             for index in range(top_word_indices.shape[0])]\n",
    "print(top_words[:3])\n",
    "print(top_word_frequency[:3])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "#### Word Cloud\n",
    "See https://github.com/amueller/word_cloud."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYEAAADKCAYAAABDsfw/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10bf05cd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from wordcloud import WordCloud\n",
    "## plot the first document\n",
    "wordcloud = WordCloud(background_color=\"white\", max_words=20)\n",
    "wordcloud.generate(corpus_stemmed_nodigits[0])\n",
    "import matplotlib.pyplot as plt\n",
    "plt.imshow(wordcloud, interpolation='bilinear')\n",
    "plt.axis(\"off\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "### 3. Run LDA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/zhiyzuo/.virtualenvs/PyEnv/lib/python2.7/site-packages/sklearn/decomposition/online_lda.py:508: DeprecationWarning: The default value for 'learning_method' will be changed from 'online' to 'batch' in the release 0.20. This warning was introduced in 0.18.\n",
      "  DeprecationWarning)\n"
     ]
    }
   ],
   "source": [
    "from sklearn.decomposition import LatentDirichletAllocation\n",
    "## large n will produce fine-grained topics\n",
    "## small n will produce coarse-grained topics\n",
    "lda = LatentDirichletAllocation(n_topics=5)\n",
    "corpus_lda = lda.fit_transform(corpus_counter)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": [
    "def print_top_words(model, feature_names, n_top_words):\n",
    "    top_word_list = []\n",
    "    for topic_idx, topic in enumerate(model.components_):\n",
    "        print(\"Topic #%d:\" % topic_idx)\n",
    "        top_words = [feature_names[i] \\\n",
    "                     for i in topic.argsort()[:-n_top_words - 1:-1]]\n",
    "        print(\" \".join(top_words))\n",
    "        top_word_list.append(top_words)\n",
    "        \n",
    "    return top_word_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Topic #0:\n",
      "class learn thi size equival data markov dimension approach set\n",
      "Topic #1:\n",
      "data label learn model function algorithm thi stream set use\n",
      "Topic #2:\n",
      "thi algorithm data cluster use problem method model learn estim\n",
      "Topic #3:\n",
      "algorithm learn problem use thi bound method function sampl model\n",
      "Topic #4:\n",
      "data model method learn thi infer use propos variat machin\n"
     ]
    }
   ],
   "source": [
    "top_word_list = print_top_words(lda, words, 10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYEAAADKCAYAAABDsfw/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10bf425d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "lda_wordcloud = WordCloud(background_color=\"white\", max_words=10, relative_scaling=0)\n",
    "lda_wordcloud.generate(' '.join(top_word_list[4]))\n",
    "import matplotlib.pyplot as plt\n",
    "plt.imshow(lda_wordcloud, interpolation='bilinear')\n",
    "plt.axis(\"off\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 0.40434032  0.3684704   0.35107134 ...,  5.23276586  0.3734939\n",
      "   1.13729281]\n",
      " [ 0.42525372  0.33952421  0.35985204 ...,  1.17029832  0.40808637\n",
      "   0.32377033]\n",
      " [ 0.35091375  1.17582242  0.36610332 ...,  3.67024033  3.03098443\n",
      "   1.15579953]\n",
      " [ 3.50968539  0.36678288  1.127191   ...,  3.32732841  2.51637274\n",
      "   0.40427492]\n",
      " [ 0.36939509  1.14205867  1.15214277 ...,  2.74948649  0.36806541\n",
      "   0.35835541]]\n",
      "(5, 789)\n"
     ]
    }
   ],
   "source": [
    "print(lda.components_)\n",
    "print(lda.components_.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "### 4. Conclusion"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "Of course there are a lot of things to do. We can, for example, compare the document similarity across the latent topic space. To interprete the meaning of the topics corresponding to the top keywords, we may need to rerun LDA with different number of topics. This tutorial provides a simple procedure on how we can conduct text analyses in Python (in this case, dimensionality reduction by mapping document and words onto a lower-dimemnsion topic space)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "While `scikit-learn` provides many handy functions to conduct topic modeling, I actually recommend using the `LDA` in package `gensim` (see <a href=\"https://radimrehurek.com/gensim/wiki.html#latent-dirichlet-allocation\" target=\"_blank\">here</a>). One of the reasons is that it is very hard to interprete what the `components_` mean in the trained `LDA` model in `sklearn`: "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "Finally, let's try to use a fantastic tool `pyLDAVis`. See <a href=\"http://nbviewer.jupyter.org/github/bmabey/pyLDAvis/blob/master/notebooks/sklearn.ipynb\" target=\"_blank\">here</a> for a more detailed example. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/Cellar/python/2.7.13/Frameworks/Python.framework/Versions/2.7/lib/python2.7/json/encoder.py:207: DeprecationWarning: Interpreting naive datetime as local 2017-04-02 18:07:47.607245. Please add timezone info to timestamps.\n",
      "  chunks = self.iterencode(o, _one_shot=True)\n"
     ]
    }
   ],
   "source": [
    "import pyLDAvis\n",
    "import pyLDAvis.sklearn\n",
    "pyLDAvis.enable_notebook()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "<link rel=\"stylesheet\" type=\"text/css\" href=\"https://cdn.rawgit.com/bmabey/pyLDAvis/files/ldavis.v1.0.0.css\">\n",
       "\n",
       "\n",
       "<div id=\"ldavis_el7836844925577765330763488\"></div>\n",
       "<script type=\"text/javascript\">\n",
       "\n",
       "var ldavis_el7836844925577765330763488_data = {\"plot.opts\": {\"xlab\": \"PC1\", \"ylab\": \"PC2\"}, \"topic.order\": [5, 3, 1, 4, 2], \"token.table\": {\"Topic\": [1, 1, 2, 3, 4, 5, 1, 1, 1, 2, 3, 4, 1, 2, 3, 4, 5, 1, 2, 3, 4, 5, 1, 2, 1, 2, 5, 4, 1, 2, 1, 4, 1, 2, 3, 5, 1, 3, 2, 3, 2, 2, 5, 3, 4, 2, 5, 2, 5, 1, 2, 3, 5, 1, 3, 4, 1, 1, 3, 4, 1, 1, 3, 1, 3, 4, 5, 1, 2, 3, 4, 5, 1, 2, 4, 1, 4, 1, 3, 4, 1, 2, 5, 3, 1, 1, 4, 1, 2, 4, 1, 1, 2, 1, 2, 3, 4, 1, 4, 1, 3, 5, 1, 2, 5, 1, 2, 1, 2, 3, 5, 5, 2, 1, 4, 1, 2, 4, 1, 3, 4, 1, 3, 2, 1, 2, 3, 1, 2, 3, 4, 5, 2, 5, 3, 5, 1, 2, 3, 4, 5, 1, 1, 2, 3, 5, 1, 2, 3, 4, 5, 3, 5, 1, 5, 1, 2, 3, 1, 2, 3, 5, 1, 3, 1, 2, 3, 4, 5, 4, 5, 4, 3, 5, 1, 1, 4, 1, 2, 3, 4, 5, 1, 2, 3, 4, 5, 1, 2, 3, 5, 2, 1, 2, 3, 4, 4, 5, 1, 2, 4, 1, 2, 3, 4, 5, 1, 2, 3, 5, 2, 3, 5, 1, 1, 3, 1, 5, 1, 4, 4, 5, 1, 1, 2, 4, 1, 2, 3, 4, 5, 5, 1, 2, 3, 4, 1, 2, 3, 1, 3, 5, 1, 3, 4, 3, 4, 1, 2, 3, 1, 4, 5, 1, 4, 1, 2, 4, 5, 2, 3, 4, 5, 1, 2, 3, 4, 5, 2, 3, 1, 2, 3, 5, 1, 1, 3, 5, 1, 2, 3, 4, 5, 1, 3, 1, 3, 2, 1, 2, 4, 1, 2, 3, 5, 1, 3, 2, 4, 1, 2, 3, 4, 5, 1, 4, 1, 2, 3, 4, 5, 1, 2, 3, 1, 2, 3, 5, 3, 5, 1, 2, 3, 4, 1, 2, 4, 1, 2, 3, 4, 5, 5, 1, 1, 1, 2, 3, 1, 3, 5, 1, 2, 3, 4, 5, 1, 3, 4, 2, 1, 3, 5, 1, 2, 3, 5, 4, 5, 1, 2, 3, 4, 5, 1, 2, 5, 1, 2, 3, 4, 5, 1, 2, 3, 1, 2, 3, 1, 2, 5, 2, 5, 1, 2, 3, 5, 1, 1, 2, 3, 4, 5, 5, 1, 2, 1, 2, 3, 5, 2, 1, 1, 2, 3, 4, 5, 1, 3, 4, 2, 3, 2, 1, 4, 1, 3, 4, 5, 1, 2, 3, 5, 1, 2, 4, 1, 2, 3, 4, 5, 1, 2, 3, 5, 1, 2, 3, 4, 5, 2, 1, 2, 1, 5, 1, 3, 4, 5, 1, 1, 2, 3, 4, 5, 1, 2, 3, 4, 5, 1, 2, 3, 4, 5, 1, 3, 1, 2, 4, 1, 5, 1, 2, 5, 5, 5, 1, 3, 4, 5, 1, 2, 5, 1, 4, 1, 2, 5, 4, 4, 1, 2, 3, 4, 1, 2, 5, 1, 2, 3, 4, 5, 2, 2, 3, 1, 2, 3, 4, 5, 2, 1, 3, 1, 5, 2, 1, 4, 1, 2, 3, 4, 5, 1, 4, 1, 3, 1, 2, 3, 4, 5, 2, 4, 1, 2, 3, 4, 1, 2, 3, 5, 2, 2, 3, 1, 2, 2, 1, 2, 1, 2, 3, 4, 1, 2, 3, 4, 5, 5, 1, 3, 2, 3, 5, 1, 2, 3, 4, 5, 1, 4, 5, 2, 3, 1, 2, 3, 4, 5, 1, 2, 5, 1, 2, 3, 5, 1, 4, 5, 1, 2, 3, 4, 1, 3, 1, 4, 1, 5, 1, 2, 3, 4, 5, 1, 2, 3, 3, 1, 3, 1, 2, 4, 3, 4, 1, 3, 1, 4], \"Freq\": [0.834752151558033, 0.5419843237368264, 0.28454176996183383, 0.09484725665394461, 0.0677480404671033, 0.02709921618684132, 0.890015035364864, 0.8294294182816525, 0.4977663463209866, 0.14221895609171045, 0.2844379121834209, 0.07110947804585523, 0.5829476932780553, 0.21198097937383828, 0.05299524484345957, 0.10599048968691914, 0.05299524484345957, 0.31387213893094046, 0.3661841620860972, 0.15693606946547023, 0.1046240463103135, 0.05231202315515675, 0.15409022896210675, 0.616360915848427, 0.28080092925065075, 0.28080092925065075, 0.28080092925065075, 0.6829402751486598, 0.18999302723467426, 0.66497559532136, 0.761660293626621, 0.169257843028138, 0.4110528185246533, 0.319707747741397, 0.1826901415665126, 0.0913450707832563, 0.22674912402717382, 0.45349824805434763, 0.6550183115599224, 0.21833943718664078, 0.629265830565862, 0.6655299548366151, 0.22184331827887172, 0.7186324292848097, 0.17965810732120244, 0.7555348813802647, 0.18888372034506618, 0.2901368565251058, 0.2901368565251058, 0.4375564877698991, 0.35800076272082654, 0.07955572504907256, 0.11933358757360885, 0.17552777078218992, 0.35105554156437985, 0.35105554156437985, 0.8289143394359989, 0.37506982746578776, 0.25004655164385853, 0.37506982746578776, 0.8346170689483422, 0.17214130723614976, 0.688565228944599, 0.24793186149038468, 0.24793186149038468, 0.24793186149038468, 0.24793186149038468, 0.8622655189684502, 0.07185545991403752, 0.02395181997134584, 0.02395181997134584, 0.02395181997134584, 0.6304256252922444, 0.04849427886863419, 0.2909656732118051, 0.4340978905817264, 0.4340978905817264, 0.3365571134663155, 0.3365571134663155, 0.16827855673315775, 0.6802570832519444, 0.25914555552455026, 0.03239319444056878, 0.6765979055578472, 0.821206589382922, 0.22891420847728822, 0.45782841695457643, 0.7537996273039206, 0.08375551414488007, 0.08375551414488007, 0.7253246717571122, 0.7336515083818179, 0.1048073583402597, 0.23446286382775822, 0.23446286382775822, 0.4103100116985769, 0.11723143191387911, 0.26729390881245074, 0.5345878176249015, 0.7727126660882914, 0.11038752372689876, 0.05519376186344938, 0.8535737286929282, 0.05021021933487813, 0.05021021933487813, 0.17873083772398374, 0.714923350895935, 0.3709523303728747, 0.24730155358191644, 0.12365077679095822, 0.24730155358191644, 0.37916128343321676, 0.7836855172627063, 0.2293972181287587, 0.4587944362575174, 0.10899594573410022, 0.21799189146820044, 0.5449797286705012, 0.3720279128276792, 0.49603721710357224, 0.12400930427589306, 0.7701907656642426, 0.1283651276107071, 0.6363352016161835, 0.22557177512594134, 0.22557177512594134, 0.4511435502518827, 0.6926400224636269, 0.18470400599030048, 0.02308800074878756, 0.08080800262075646, 0.03463200112318134, 0.6305989132954936, 0.37511891427719196, 0.6277738278577423, 0.2092579426192474, 0.4561828455039373, 0.3193279918527561, 0.09123656910078747, 0.09123656910078747, 0.09123656910078747, 0.8679971578628688, 0.15788417491478643, 0.4736525247443593, 0.15788417491478643, 0.3731105233244858, 0.35468037965122995, 0.23645358643415332, 0.039408931072358884, 0.23645358643415332, 0.15763572428943554, 0.48120143000299187, 0.24060071500149594, 0.20181038490233605, 0.6054311547070081, 0.8592581289524354, 0.27168463027829143, 0.5433692605565829, 0.15453534174495873, 0.15453534174495873, 0.46360602523487615, 0.15453534174495873, 0.342398312243017, 0.342398312243017, 0.45527281968694416, 0.2731636918121665, 0.09105456393738882, 0.13658184590608324, 0.04552728196869441, 0.3670489831960329, 0.3670489831960329, 0.3587117430619631, 0.16295060869662503, 0.6518024347865001, 0.8190319844968682, 0.34608797032230304, 0.34608797032230304, 0.6034066501977564, 0.1149346000376679, 0.1149346000376679, 0.08620095002825093, 0.08620095002825093, 0.361439717585329, 0.3098054722159963, 0.20653698147733085, 0.10326849073866542, 0.05163424536933271, 0.08064682085823158, 0.24194046257469473, 0.564527746007621, 0.08064682085823158, 0.6351339287264428, 0.6761298322983416, 0.09658997604262022, 0.09658997604262022, 0.09658997604262022, 0.36083579847239355, 0.36083579847239355, 0.3790847409010831, 0.5212415187389893, 0.09477118522527077, 0.507422102373091, 0.10148442047461823, 0.12685552559327276, 0.228339946067891, 0.07611331535596366, 0.4732319153904193, 0.23661595769520966, 0.15774397179680644, 0.15774397179680644, 0.23942977520384, 0.23942977520384, 0.23942977520384, 0.9438239141306088, 0.9034165401677423, 0.08212877637888567, 0.28031281810746866, 0.28031281810746866, 0.3449304014612883, 0.3449304014612883, 0.30562414757448386, 0.30562414757448386, 0.7997855720917364, 0.624132636946784, 0.156033159236696, 0.23404973885504401, 0.4843412015880551, 0.20757480068059503, 0.06919160022686502, 0.06919160022686502, 0.13838320045373004, 0.378404697894677, 0.48185302836513055, 0.2810809332129928, 0.20077209515213773, 0.040154419030427546, 0.7953892832937124, 0.05302595221958083, 0.10605190443916165, 0.2547261753456174, 0.5094523506912348, 0.1273630876728087, 0.22642058896228856, 0.22642058896228856, 0.4528411779245771, 0.5697894302167913, 0.28489471510839565, 0.7160006241967789, 0.15342870518502405, 0.1022858034566827, 0.26710011784135945, 0.5342002356827189, 0.3700386635358213, 0.2587266465706767, 0.6468166164266917, 0.6658933072050601, 0.14797629049001335, 0.14797629049001335, 0.036994072622503336, 0.6449821757977606, 0.6605799340243969, 0.1651449835060992, 0.1651449835060992, 0.4252809525545205, 0.30567068464856156, 0.02658005953465753, 0.13290029767328765, 0.10632023813863012, 0.19719394353078756, 0.5915818305923627, 0.44684939613401764, 0.22342469806700882, 0.11171234903350441, 0.22342469806700882, 0.8926668023700137, 0.27937357800347556, 0.46562263000579257, 0.09312452600115852, 0.4825663795337757, 0.14476991386013272, 0.24128318976688784, 0.04825663795337757, 0.09651327590675514, 0.34580888427744716, 0.34580888427744716, 0.0956967525868723, 0.7655740206949784, 0.6315016736614035, 0.41742268117525516, 0.5009072174103062, 0.041742268117525516, 0.7925986610400269, 0.03302494421000112, 0.06604988842000224, 0.06604988842000224, 0.3395143370074896, 0.3395143370074896, 0.6629904818144943, 0.22099682727149808, 0.5949078667667762, 0.04576214359744432, 0.27457286158466593, 0.04576214359744432, 0.09152428719488864, 0.268038768527255, 0.268038768527255, 0.4361875383199346, 0.25444273068662854, 0.2180937691599673, 0.07269792305332244, 0.03634896152666122, 0.5280023331090702, 0.10560046662181403, 0.3168013998654421, 0.21285868709395425, 0.21285868709395425, 0.21285868709395425, 0.4257173741879085, 0.2997804009304328, 0.2997804009304328, 0.3369609671710368, 0.11232032239034559, 0.22464064478069118, 0.3369609671710368, 0.5277271766380839, 0.19789769123928144, 0.26386358831904194, 0.6455853426222337, 0.13591270370994393, 0.11892361574620093, 0.05096726389122897, 0.06795635185497197, 0.6159818092629549, 0.8660061482068998, 0.867270475354425, 0.10685857075094321, 0.6411514245056593, 0.21371714150188642, 0.255039341570808, 0.255039341570808, 0.510078683141616, 0.6031115853227551, 0.17652046399690394, 0.11768030933126929, 0.05884015466563464, 0.044130115999225984, 0.25304284694722234, 0.25304284694722234, 0.3795642704208335, 0.6161051745532372, 0.26407633282634124, 0.26407633282634124, 0.26407633282634124, 0.6601222082293929, 0.20628819007168528, 0.08251527602867412, 0.04125763801433706, 0.5891181970131393, 0.29455909850656964, 0.4555307722783845, 0.24847133033366428, 0.041411888388944045, 0.20705944194472023, 0.041411888388944045, 0.4408081335694993, 0.4408081335694993, 0.07346802226158322, 0.42274834293064417, 0.3623557225119807, 0.06039262041866345, 0.06039262041866345, 0.06039262041866345, 0.5406249778911779, 0.3964583171201972, 0.036041665192745195, 0.7413873504152572, 0.1853468376038143, 0.061782279201271426, 0.142692946521753, 0.42807883956525894, 0.285385893043506, 0.2856827030122556, 0.2856827030122556, 0.6350126066735747, 0.16710858070357232, 0.06684343228142893, 0.13368686456285786, 0.8297810108752663, 0.6545251484175553, 0.0981787722626333, 0.0981787722626333, 0.06545251484175553, 0.06545251484175553, 0.3884765002104408, 0.20664382263300363, 0.6199314678990109, 0.3662859209659219, 0.12209530698864063, 0.3662859209659219, 0.12209530698864063, 0.750792131453917, 0.8321945462572043, 0.5998502286397716, 0.15996006097060575, 0.03999001524265144, 0.07998003048530287, 0.07998003048530287, 0.17861334415207927, 0.4465333603801982, 0.35722668830415855, 0.2299847229489034, 0.4599694458978068, 0.6281817406881598, 0.4320306816014703, 0.4320306816014703, 0.31015921454441686, 0.20677280969627793, 0.31015921454441686, 0.20677280969627793, 0.4418628283778342, 0.2761642677361464, 0.2209314141889171, 0.055232853547229274, 0.43268913252770763, 0.21634456626385382, 0.3245168493957807, 0.6719302258250541, 0.18951878164296398, 0.034457960298720725, 0.05168694044808109, 0.05168694044808109, 0.19684305203828112, 0.4921076300957028, 0.19684305203828112, 0.3670938359567438, 0.41871430847061686, 0.08374286169412337, 0.3349714467764935, 0.08374286169412337, 0.08374286169412337, 0.8342756033572684, 0.17893067748683558, 0.5367920324605068, 0.7518269015178588, 0.10740384307397983, 0.5313361802928052, 0.15181033722651577, 0.15181033722651577, 0.22771550583977365, 0.8297520865760214, 0.6507053327477595, 0.11619738084781421, 0.06971842850868852, 0.13943685701737704, 0.02323947616956284, 0.38089933950414445, 0.25393289300276295, 0.21161074416896913, 0.08464429766758766, 0.08464429766758766, 0.5724883172672505, 0.2671612147247169, 0.07633177563563338, 0.07633177563563338, 0.03816588781781669, 0.8250165320260913, 0.08250165320260913, 0.7450730629615854, 0.1241788438269309, 0.1241788438269309, 0.6538343185440622, 0.2615337274176249, 0.44258937831719786, 0.14752979277239928, 0.29505958554479855, 0.3838527003511136, 0.3627842530769966, 0.39650671853321634, 0.09912667963330409, 0.09912667963330409, 0.39650671853321634, 0.2541391670657965, 0.508278334131593, 0.12706958353289824, 0.26745714234702933, 0.26745714234702933, 0.3417240115876365, 0.17086200579381824, 0.3417240115876365, 0.5678735124707827, 0.6999277613560086, 0.2370221309200162, 0.31602950789335493, 0.2370221309200162, 0.15801475394667747, 0.22982181686715983, 0.22982181686715983, 0.22982181686715983, 0.608470773100352, 0.11408826995631599, 0.076058846637544, 0.11408826995631599, 0.076058846637544, 0.6271019630864143, 0.7756740390608232, 0.1292790065101372, 0.5310156830769557, 0.14750435641026546, 0.20650609897437164, 0.08850261384615928, 0.02950087128205309, 0.6426587382240775, 0.2636680627302006, 0.5273361254604012, 0.8399644522664995, 0.11999492175235707, 0.6231595310901745, 0.3739483252319856, 0.4985977669759808, 0.6623714386341426, 0.19319166960162493, 0.027598809943089275, 0.08279642982926783, 0.027598809943089275, 0.26754612783204323, 0.5350922556640865, 0.3413611118995199, 0.3413611118995199, 0.7829306149636107, 0.09786632687045134, 0.04893316343522567, 0.04893316343522567, 0.04893316343522567, 0.39252532173303095, 0.39252532173303095, 0.3307590374302698, 0.3307590374302698, 0.13230361497210794, 0.26460722994421587, 0.5069884176074039, 0.16899613920246795, 0.16899613920246795, 0.1126640928016453, 0.758486016183052, 0.3845082688136738, 0.3845082688136738, 0.13812652210089313, 0.6906326105044656, 0.756492446369252, 0.24448511185444535, 0.6112127796361133, 0.14016671659799335, 0.7008335829899668, 0.14016671659799335, 0.5686333928740533, 0.6264565380030428, 0.09637792892354505, 0.09637792892354505, 0.14456689338531759, 0.048188964461772524, 0.3725858139975918, 0.8883422699665632, 0.06345301928332595, 0.5676620842369324, 0.22706483369477296, 0.11353241684738648, 0.33249909409550116, 0.2216660627303341, 0.055416515682583524, 0.2770825784129176, 0.055416515682583524, 0.6450788213284347, 0.08063485266605434, 0.24190455799816304, 0.27236087032117884, 0.5447217406423577, 0.679876959825172, 0.15505965750398662, 0.0477106638473805, 0.07156599577107074, 0.0477106638473805, 0.21112299627049075, 0.21112299627049075, 0.4222459925409815, 0.37241963194954864, 0.26601402282110614, 0.21281121825688493, 0.1596084136926637, 0.1652958435906201, 0.49588753077186026, 0.1652958435906201, 0.6775031114320339, 0.09678615877600484, 0.09678615877600484, 0.09678615877600484, 0.23502122182979351, 0.7050636654893805, 0.7253956371726197, 0.14507912743452392, 0.21741610323477406, 0.4348322064695481, 0.5635863238104566, 0.23729950476229753, 0.08898731428586157, 0.07415609523821798, 0.02966243809528719, 0.1424762308260159, 0.4274286924780477, 0.4274286924780477, 0.689068225377359, 0.37523032163268755, 0.37523032163268755, 0.7663040062175928, 0.7365786376322627, 0.18414465940806568, 0.28408324557840187, 0.5681664911568037, 0.4394149952818145, 0.4394149952818145, 0.2657826261643013, 0.5315652523286026], \"Term\": [\"algebra\", \"algorithm\", \"algorithm\", \"algorithm\", \"algorithm\", \"algorithm\", \"ani\", \"anomali\", \"appli\", \"appli\", \"appli\", \"appli\", \"approach\", \"approach\", \"approach\", \"approach\", \"approach\", \"approxim\", \"approxim\", \"approxim\", \"approxim\", \"approxim\", \"arm\", \"arm\", \"asymptot\", \"asymptot\", \"asymptot\", \"attribut\", \"bandit\", \"bandit\", \"bas\", \"bas\", \"bayesian\", \"bayesian\", \"bayesian\", \"bayesian\", \"becaus\", \"becaus\", \"behavior\", \"behavior\", \"belong\", \"best\", \"best\", \"binari\", \"binari\", \"block\", \"block\", \"boolean\", \"boolean\", \"bound\", \"bound\", \"bound\", \"bound\", \"box\", \"box\", \"box\", \"captur\", \"carlo\", \"carlo\", \"carlo\", \"causal\", \"chain\", \"chain\", \"character\", \"character\", \"character\", \"character\", \"class\", \"class\", \"class\", \"class\", \"class\", \"classif\", \"classif\", \"classif\", \"classification\", \"classification\", \"close\", \"close\", \"close\", \"cluster\", \"cluster\", \"cluster\", \"column\", \"combinatori\", \"commun\", \"commun\", \"compar\", \"compar\", \"compar\", \"comparison\", \"complet\", \"complet\", \"complex\", \"complex\", \"complex\", \"complex\", \"compon\", \"compon\", \"condit\", \"condit\", \"condit\", \"consist\", \"consist\", \"consist\", \"constant\", \"constant\", \"constraint\", \"constraint\", \"constraint\", \"constraint\", \"contain\", \"contextu\", \"correctli\", \"correctli\", \"cost\", \"cost\", \"cost\", \"count\", \"count\", \"count\", \"criterion\", \"criterion\", \"critic\", \"cross\", \"cross\", \"cross\", \"data\", \"data\", \"data\", \"data\", \"data\", \"decade\", \"designs\", \"dimens\", \"dimens\", \"dimension\", \"dimension\", \"dimension\", \"dimension\", \"dimension\", \"direct\", \"discret\", \"discret\", \"discret\", \"discuss\", \"distribut\", \"distribut\", \"distribut\", \"distribut\", \"distribut\", \"distributions\", \"distributions\", \"divers\", \"divers\", \"domain\", \"dure\", \"dure\", \"dynam\", \"dynam\", \"dynam\", \"dynam\", \"easili\", \"easili\", \"effici\", \"effici\", \"effici\", \"effici\", \"effici\", \"efficient\", \"efficient\", \"embed\", \"ensembl\", \"ensembl\", \"entri\", \"especi\", \"especi\", \"estim\", \"estim\", \"estim\", \"estim\", \"estim\", \"exist\", \"exist\", \"exist\", \"exist\", \"exist\", \"factor\", \"factor\", \"factor\", \"factor\", \"factorization\", \"featur\", \"featur\", \"featur\", \"featur\", \"fourier\", \"fourier\", \"framework\", \"framework\", \"framework\", \"function\", \"function\", \"function\", \"function\", \"function\", \"gener\", \"gener\", \"gener\", \"gener\", \"global\", \"global\", \"global\", \"graph\", \"graphic\", \"graphic\", \"greedi\", \"greedi\", \"hilbert\", \"hilbert\", \"homotopi\", \"homotopi\", \"imbalanc\", \"implement\", \"implement\", \"implement\", \"improv\", \"improv\", \"improv\", \"improv\", \"improv\", \"include\", \"infer\", \"infer\", \"infer\", \"infer\", \"inform\", \"inform\", \"inform\", \"information\", \"information\", \"information\", \"integr\", \"integr\", \"integr\", \"integrand\", \"integrand\", \"introduc\", \"introduc\", \"introduc\", \"invari\", \"invari\", \"journal\", \"kernel\", \"kernel\", \"label\", \"label\", \"label\", \"label\", \"largest\", \"latent\", \"latent\", \"latent\", \"learn\", \"learn\", \"learn\", \"learn\", \"learn\", \"learner\", \"learner\", \"learning\", \"learning\", \"learning\", \"learning\", \"level\", \"likelihood\", \"likelihood\", \"likelihood\", \"linear\", \"linear\", \"linear\", \"linear\", \"linear\", \"littl\", \"littl\", \"log\", \"log\", \"logarithm\", \"low\", \"low\", \"low\", \"machin\", \"machin\", \"machin\", \"machin\", \"maintain\", \"maintain\", \"manifold\", \"manifold\", \"markov\", \"markov\", \"markov\", \"markov\", \"markov\", \"matlab\", \"matlab\", \"matrix\", \"matrix\", \"matrix\", \"matrix\", \"matrix\", \"maxim\", \"maxim\", \"maxim\", \"maximum\", \"maximum\", \"maximum\", \"maximum\", \"mcmc\", \"mcmc\", \"mean\", \"mean\", \"mean\", \"mean\", \"measur\", \"measur\", \"measur\", \"method\", \"method\", \"method\", \"method\", \"method\", \"minima\", \"minor\", \"miss\", \"mixtur\", \"mixtur\", \"mixtur\", \"mle\", \"mle\", \"mle\", \"model\", \"model\", \"model\", \"model\", \"model\", \"mont\", \"mont\", \"mont\", \"motiv\", \"mrf\", \"mrf\", \"mrf\", \"multi\", \"multi\", \"multi\", \"multi\", \"neural\", \"neural\", \"new\", \"new\", \"new\", \"new\", \"new\", \"non\", \"non\", \"non\", \"novel\", \"novel\", \"novel\", \"novel\", \"novel\", \"number\", \"number\", \"number\", \"onli\", \"onli\", \"onli\", \"onlin\", \"onlin\", \"onlin\", \"open\", \"open\", \"optim\", \"optim\", \"optim\", \"optim\", \"ord\", \"paper\", \"paper\", \"paper\", \"paper\", \"paper\", \"papers\", \"partial\", \"partial\", \"partit\", \"partit\", \"partit\", \"partit\", \"past\", \"penal\", \"perform\", \"perform\", \"perform\", \"perform\", \"perform\", \"point\", \"point\", \"point\", \"poisson\", \"poisson\", \"polici\", \"possibl\", \"possibl\", \"predict\", \"predict\", \"predict\", \"predict\", \"present\", \"present\", \"present\", \"present\", \"probabl\", \"probabl\", \"probabl\", \"problem\", \"problem\", \"problem\", \"problem\", \"problem\", \"procedur\", \"procedur\", \"procedur\", \"proceed\", \"process\", \"process\", \"process\", \"process\", \"process\", \"processes\", \"product\", \"product\", \"program\", \"program\", \"properti\", \"properti\", \"properti\", \"properti\", \"proport\", \"propos\", \"propos\", \"propos\", \"propos\", \"propos\", \"random\", \"random\", \"random\", \"random\", \"random\", \"rank\", \"rank\", \"rank\", \"rank\", \"rank\", \"rate\", \"rate\", \"real\", \"real\", \"real\", \"recov\", \"recov\", \"recoveri\", \"recoveri\", \"recoveri\", \"recovery\", \"regim\", \"regress\", \"regress\", \"regress\", \"regress\", \"regret\", \"regret\", \"regret\", \"relationship\", \"relationship\", \"relax\", \"relax\", \"relax\", \"relev\", \"reproduc\", \"requir\", \"requir\", \"requir\", \"requir\", \"research\", \"research\", \"research\", \"result\", \"result\", \"result\", \"result\", \"result\", \"retain\", \"robust\", \"robust\", \"sampl\", \"sampl\", \"sampl\", \"sampl\", \"sampl\", \"samples\", \"scalabl\", \"scalabl\", \"second\", \"second\", \"sens\", \"sensit\", \"sensit\", \"set\", \"set\", \"set\", \"set\", \"set\", \"shift\", \"shift\", \"signific\", \"signific\", \"size\", \"size\", \"size\", \"size\", \"size\", \"sketch\", \"sketch\", \"space\", \"space\", \"space\", \"space\", \"spars\", \"spars\", \"spars\", \"spars\", \"sparsiti\", \"stochast\", \"stochast\", \"stream\", \"stream\", \"strong\", \"structur\", \"structur\", \"sup\", \"sup\", \"sup\", \"supervis\", \"support\", \"support\", \"support\", \"support\", \"support\", \"surrog\", \"svm\", \"svm\", \"tensor\", \"tensor\", \"tensor\", \"theoret\", \"theoret\", \"theoret\", \"theoret\", \"theoret\", \"theori\", \"theori\", \"theori\", \"therebi\", \"therebi\", \"thi\", \"thi\", \"thi\", \"thi\", \"thi\", \"tight\", \"tight\", \"tight\", \"time\", \"time\", \"time\", \"time\", \"topic\", \"topic\", \"topic\", \"train\", \"train\", \"train\", \"train\", \"transit\", \"transit\", \"tree\", \"tree\", \"uniform\", \"uniform\", \"use\", \"use\", \"use\", \"use\", \"use\", \"variat\", \"variat\", \"variat\", \"vectors\", \"version\", \"version\", \"video\", \"weak\", \"weak\", \"web\", \"web\", \"wide\", \"wide\", \"word\", \"word\"]}, \"mdsDat\": {\"y\": [0.014841103350945167, 0.07192418607236398, -0.012134291141910914, -0.053080076220073436, -0.021550922061324803], \"cluster\": [1, 1, 1, 1, 1], \"Freq\": [49.93191929905031, 20.78200044285792, 11.774719237061673, 9.973945800501893, 7.537415220528206], \"topics\": [1, 2, 3, 4, 5], \"x\": [0.03367342140537426, 0.011931666624606332, -0.08121295072700409, 0.049985655962883534, -0.014377793265860023]}, \"R\": 30, \"lambda.step\": 0.01, \"tinfo\": {\"Category\": [\"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\"], \"Term\": [\"learn\", \"kernel\", \"function\", \"distribut\", \"classif\", \"log\", \"factor\", \"markov\", \"regress\", \"gener\", \"variat\", \"point\", \"cost\", \"theoret\", \"complex\", \"optim\", \"likelihood\", \"bound\", \"ensembl\", \"sampl\", \"time\", \"properti\", \"theori\", \"linear\", \"data\", \"low\", \"sensit\", \"method\", \"framework\", \"new\", \"graph\", \"svm\", \"ord\", \"consist\", \"graphic\", \"class\", \"imbalanc\", \"rate\", \"level\", \"ani\", \"minor\", \"miss\", \"criterion\", \"domain\", \"video\", \"inform\", \"machin\", \"proport\", \"combinatori\", \"algebra\", \"direct\", \"program\", \"second\", \"causal\", \"anomali\", \"entri\", \"tree\", \"penal\", \"comparison\", \"captur\", \"size\", \"condit\", \"bas\", \"complet\", \"data\", \"thi\", \"problem\", \"compar\", \"real\", \"onli\", \"method\", \"cluster\", \"model\", \"set\", \"propos\", \"introduc\", \"train\", \"label\", \"use\", \"paper\", \"optim\", \"algorithm\", \"featur\", \"estim\", \"multi\", \"approach\", \"result\", \"learn\", \"perform\", \"function\", \"sampl\", \"number\", \"rank\", \"contextu\", \"robust\", \"sup\", \"processes\", \"polici\", \"stream\", \"bandit\", \"sparsiti\", \"past\", \"strong\", \"block\", \"constant\", \"arm\", \"weak\", \"manifold\", \"structur\", \"decade\", \"factorization\", \"largest\", \"belong\", \"sens\", \"mixtur\", \"logarithm\", \"samples\", \"critic\", \"behavior\", \"product\", \"motiv\", \"retain\", \"best\", \"tensor\", \"low\", \"framework\", \"regret\", \"procedur\", \"partial\", \"discret\", \"learn\", \"number\", \"non\", \"algorithm\", \"variat\", \"bound\", \"use\", \"approxim\", \"data\", \"cluster\", \"novel\", \"model\", \"problem\", \"thi\", \"rank\", \"dimension\", \"bayesian\", \"approach\", \"infer\", \"matrix\", \"method\", \"set\", \"new\", \"effici\", \"random\", \"gener\", \"log\", \"transit\", \"binari\", \"chain\", \"dimens\", \"factor\", \"latent\", \"learner\", \"vectors\", \"column\", \"distributions\", \"count\", \"information\", \"poisson\", \"littl\", \"easili\", \"becaus\", \"likelihood\", \"signific\", \"maintain\", \"dynam\", \"integrand\", \"wide\", \"therebi\", \"point\", \"scalabl\", \"stochast\", \"cross\", \"dure\", \"box\", \"variat\", \"complex\", \"process\", \"markov\", \"version\", \"close\", \"appli\", \"linear\", \"partit\", \"sampl\", \"matrix\", \"random\", \"infer\", \"model\", \"maxim\", \"present\", \"method\", \"time\", \"algorithm\", \"requir\", \"exist\", \"use\", \"function\", \"estim\", \"thi\", \"bayesian\", \"gener\", \"kernel\", \"relev\", \"supervis\", \"cost\", \"reproduc\", \"attribut\", \"embed\", \"hilbert\", \"especi\", \"commun\", \"sensit\", \"correctli\", \"topic\", \"neural\", \"web\", \"sketch\", \"invari\", \"compon\", \"word\", \"shift\", \"matlab\", \"relationship\", \"mont\", \"carlo\", \"integr\", \"mean\", \"possibl\", \"efficient\", \"classification\", \"fourier\", \"point\", \"classif\", \"theoret\", \"probabl\", \"function\", \"distribut\", \"space\", \"learn\", \"predict\", \"measur\", \"new\", \"propos\", \"implement\", \"data\", \"label\", \"support\", \"thi\", \"use\", \"approach\", \"algorithm\", \"set\", \"model\", \"estim\", \"problem\", \"method\", \"effici\", \"sampl\", \"divers\", \"ensembl\", \"minima\", \"uniform\", \"papers\", \"contain\", \"include\", \"surrog\", \"recovery\", \"designs\", \"journal\", \"discuss\", \"proceed\", \"tight\", \"mle\", \"regim\", \"regress\", \"homotopi\", \"mcmc\", \"boolean\", \"maximum\", \"greedi\", \"open\", \"onlin\", \"asymptot\", \"global\", \"mrf\", \"research\", \"relax\", \"efficient\", \"character\", \"recoveri\", \"properti\", \"distribut\", \"learn\", \"gener\", \"theori\", \"learning\", \"optim\", \"recov\", \"bound\", \"improv\", \"time\", \"method\", \"spars\", \"constraint\", \"thi\", \"function\", \"problem\", \"estim\", \"model\", \"data\", \"machin\", \"algorithm\", \"result\", \"paper\"], \"loglift\": [30.0, 29.0, 28.0, 27.0, 26.0, 25.0, 24.0, 23.0, 22.0, 21.0, 20.0, 19.0, 18.0, 17.0, 16.0, 15.0, 14.0, 13.0, 12.0, 11.0, 10.0, 9.0, 8.0, 7.0, 6.0, 5.0, 4.0, 3.0, 2.0, 1.0, 0.6143, 0.562, 0.5584, 0.5577, 0.5519, 0.5473, 0.5392, 0.5258, 0.5206, 0.5153, 0.5095, 0.5022, 0.4937, 0.4918, 0.4905, 0.4886, 0.4811, 0.475, 0.4739, 0.4719, 0.469, 0.4659, 0.4658, 0.4653, 0.4615, 0.4608, 0.4605, 0.4599, 0.4595, 0.4542, 0.4429, 0.4384, 0.4455, 0.4424, 0.323, 0.3052, 0.3076, 0.422, 0.3882, 0.3837, 0.2507, 0.3152, 0.1925, 0.2738, 0.247, 0.3474, 0.3313, 0.2845, 0.1213, 0.2606, 0.2593, 0.0739, 0.3268, 0.1958, 0.2727, 0.1369, 0.2239, -0.1752, 0.213, -0.0062, 0.0458, 0.0959, 0.1029, 1.3396, 1.2983, 1.274, 1.2663, 1.2487, 1.2306, 1.2293, 1.2067, 1.1803, 1.1734, 1.1682, 1.1611, 1.1601, 1.15, 1.1464, 1.0994, 1.0989, 1.0908, 1.0869, 1.0819, 1.0803, 1.0799, 1.0764, 1.0728, 1.0707, 1.0576, 1.0543, 1.0539, 1.0448, 1.0177, 0.9919, 0.9005, 0.8873, 0.9648, 0.8809, 0.9989, 0.9382, 0.405, 0.6006, 0.7635, 0.2943, 0.7069, 0.5323, 0.1566, 0.5209, -0.1434, 0.2722, 0.5508, -0.1569, -0.1013, -0.3191, 0.2904, 0.3869, 0.3651, 0.0207, 0.2664, 0.1528, -0.4457, -0.1356, 0.1965, 0.2665, 0.1921, 0.1144, 1.8901, 1.7499, 1.6864, 1.6542, 1.642, 1.6223, 1.5936, 1.5872, 1.5607, 1.5482, 1.5239, 1.5044, 1.5043, 1.5019, 1.4758, 1.4721, 1.4614, 1.4599, 1.4395, 1.4235, 1.3678, 1.3579, 1.3253, 1.3059, 1.2972, 1.2941, 1.293, 1.2858, 1.2834, 1.2701, 1.2608, 1.1964, 1.1118, 0.8739, 1.2391, 1.2015, 0.9171, 0.774, 1.0894, 0.5836, 0.6181, 0.588, 0.5327, 0.0342, 0.9879, 0.6264, -0.0561, 0.591, -0.2447, 0.8165, 0.5252, -0.3072, -0.0111, 0.0546, -0.8006, 0.3239, 0.1756, 1.8772, 1.8168, 1.8078, 1.7595, 1.6787, 1.6698, 1.6671, 1.6323, 1.6236, 1.6181, 1.613, 1.5874, 1.543, 1.5158, 1.4612, 1.4502, 1.4232, 1.4114, 1.409, 1.4033, 1.3899, 1.3818, 1.2546, 1.2499, 1.2395, 1.2019, 1.1989, 1.1922, 1.1912, 1.1754, 1.1524, 1.077, 1.0566, 1.1109, 0.7753, 0.7946, 0.9123, 0.3253, 1.0334, 0.8516, 0.6287, 0.2989, 0.8081, -0.2356, 0.3774, 0.5134, -0.3305, -0.204, 0.0952, -0.4765, -0.0429, -0.6334, -0.0824, -0.6425, -0.6727, 0.2642, -0.1652, 2.2159, 2.1588, 2.0426, 1.9431, 1.9193, 1.9059, 1.9005, 1.8971, 1.8938, 1.8863, 1.883, 1.8829, 1.8727, 1.8331, 1.8107, 1.8026, 1.6968, 1.6723, 1.6579, 1.6361, 1.6346, 1.6173, 1.581, 1.5231, 1.5183, 1.4766, 1.4732, 1.4198, 1.3752, 1.36, 1.3452, 1.2695, 1.09, 0.8449, 0.3574, 0.6658, 0.996, 1.1333, 0.4619, 1.2003, 0.5012, 0.8026, 0.6215, -0.1775, 0.6128, 1.1421, -0.4744, 0.0211, -0.3777, -0.0282, -0.5785, -0.8234, -0.0954, -0.9444, 0.0282, -0.1211], \"Freq\": [75.0, 15.0, 39.0, 25.0, 20.0, 10.0, 12.0, 21.0, 10.0, 25.0, 14.0, 11.0, 9.0, 18.0, 17.0, 29.0, 10.0, 25.0, 6.0, 33.0, 18.0, 13.0, 12.0, 20.0, 86.0, 23.0, 8.0, 58.0, 21.0, 24.0, 19.557238762934343, 13.80388340178982, 8.414296367614357, 17.368895562779535, 10.557670577027523, 36.03600567847296, 7.493234830132567, 10.23895048170242, 6.589892107392628, 6.574415260536, 5.758245743527523, 5.707782447474264, 6.372887026683204, 5.701818537007184, 6.384791657692578, 15.348798076677348, 24.461488590037682, 4.83803897495359, 4.883293048947266, 4.794507330251941, 5.516843545053084, 7.4082458907437365, 6.6300774279401695, 4.763438597263051, 4.775295028139444, 4.832586808912954, 5.454512069575684, 4.751960922657092, 5.449754157347587, 4.743234534605442, 15.890665939732159, 14.023961148656088, 9.212074211774928, 7.414953273532443, 59.744098488033465, 56.80191642636781, 39.41848660426719, 9.091213951445175, 11.85689360389114, 11.861884605007138, 37.76477641519174, 21.125357540498605, 41.14964930327982, 23.790767014941345, 27.504842954551762, 13.818515812221484, 14.370190084225522, 17.938570731568884, 38.006757603848776, 19.799241170294483, 19.362232851486112, 39.677494102286275, 14.335193211614225, 21.13498920153253, 15.897394858216899, 21.607971197197404, 16.424567624310143, 31.5323017976517, 15.450626480823068, 19.559900343871643, 17.718003156828367, 15.248979220214654, 14.501204852121905, 5.061494485165957, 5.888674732257309, 5.3007496893872865, 3.535108555476589, 3.459523066911449, 5.150355679078541, 7.479185725763073, 2.747430403738249, 2.7031875837489614, 2.6643895579973362, 3.538582057523574, 3.7131085338237866, 4.302503901627514, 3.5641706580603003, 5.918544861346589, 5.1043835576833265, 1.9779579905304032, 1.9479764854507058, 1.910744027366758, 1.9486260061420304, 1.9646655194787577, 5.7260659890333425, 1.931199625788254, 1.8909011188645888, 1.9056504051076328, 2.7408571484222937, 3.333385269843995, 1.9354494333527927, 1.8841760702441146, 2.591816934587876, 4.935682721604847, 12.251829700811498, 10.651489630448054, 4.292002678950945, 5.0953247567623565, 2.730629786820101, 3.3636517450352583, 23.44469037748828, 10.512702644272824, 6.069704188396129, 20.586885064424358, 5.915208129366331, 8.89688950766674, 16.387891918088325, 6.6880620375519735, 15.597673762751457, 8.422399971228154, 5.96928043134041, 12.076376469648869, 10.900652603923302, 12.662973370836081, 7.279889662717275, 6.707852557549002, 6.555222966650446, 8.006698964138664, 6.755341287256006, 6.661459841837404, 7.83311922889774, 6.575182453641872, 6.10802092974929, 5.958824665512204, 5.9502074032861705, 5.908526072346556, 8.145108466982533, 2.8827044804841138, 3.5392773105281328, 3.5764634208692248, 2.9066386449130177, 7.3948322670923465, 3.5089241896440373, 2.919786071397193, 1.6275380338089807, 1.6369658686410498, 2.246396420206466, 4.273963804369883, 4.160988605187861, 2.2988935526581513, 1.489445751520459, 1.4987754810946328, 2.2391965029921916, 5.443680572927962, 1.4551240387450874, 1.4397636749498943, 2.9917968005388262, 1.606861623592663, 2.01695461684517, 1.5956336925210801, 4.824176705562463, 1.6289508688614527, 2.2317146868139135, 1.8883291607033967, 1.5640520648770395, 2.3890187964993244, 5.831987363200157, 6.645330384845355, 4.274154837924488, 6.165562373956225, 2.166854267710633, 2.326587567108523, 4.14298917620001, 5.29113521439623, 2.866595643759762, 7.154553513649215, 6.010022224309872, 5.008754597200384, 4.995390404386124, 8.28275056495689, 2.9943645480956578, 3.9882194326479277, 6.552605661061933, 3.996501970495243, 6.804049912740982, 3.372114704367044, 3.8555526541831475, 5.839483439386776, 4.589833925128721, 4.327953508567991, 4.432917855082034, 3.564233190482547, 3.558827061579446, 10.077447729485524, 2.160995486078561, 2.138892118707221, 5.3162596551017876, 1.5272140816439557, 1.551342702552721, 1.4727355328043308, 1.4792970046763765, 1.4614675123656877, 2.1973939614236913, 4.0151238617268366, 2.1265142978900564, 2.8230991905825698, 1.541636554349688, 1.5135839792143237, 2.1670027538470626, 1.5498330115114374, 1.5305376227450274, 1.5355185734780306, 1.5168195401874291, 1.4938651532031224, 1.484916044823312, 2.7643160557666593, 2.784213319811932, 1.5214800640133792, 2.95396125262586, 1.531367974853627, 0.8951532854459663, 1.5123380676363094, 0.8954516384088832, 3.53553560770272, 6.038089026566769, 5.177436518373268, 2.8004244462881283, 8.53552994610107, 5.6023268462896745, 3.754257095017231, 10.390429847898892, 2.7115379252897975, 3.5432124670331677, 4.516159077843416, 5.787052920124172, 2.868338160828874, 6.8264640875715195, 3.9323896767423707, 3.4584783754994004, 6.008779801127536, 5.484071679288529, 4.140107381795838, 4.571030494438291, 3.4620874651730302, 3.598906421337502, 3.1967079797345814, 3.0449732147656268, 2.9958962383713317, 2.8531059459202983, 2.8659847555136886, 3.4247988961404343, 4.006240129721764, 1.8869716909872667, 2.419923352558324, 1.3224677417477821, 1.3369437961763637, 1.3324905855697127, 1.3485946253578074, 1.3047526392976463, 1.3251121897681273, 1.338858343388784, 1.3278064769529685, 1.3357729031666452, 2.2325964268500584, 1.8071216011605375, 1.2601369914263738, 4.148944676348547, 1.3131870673446744, 1.3195311911004248, 1.3339733727395124, 1.8156471494077129, 1.3550095870419199, 1.2822331307134214, 2.422560635314591, 2.4504549717866904, 1.3781702339567241, 2.4906971568509078, 1.3565694637419, 1.7451212566349572, 0.8000993606092116, 1.1671006924994007, 1.818279998790241, 2.9533379578460486, 4.452019431350934, 8.108164484579165, 3.7196170918079425, 2.530862948684227, 2.095698806639986, 3.579213154047534, 1.914364706803672, 3.127734623466117, 2.4307418559997935, 2.6376017875787587, 3.7149140309499575, 2.4695706184152, 1.9100622567599181, 3.9321118527923953, 3.0342845413048143, 2.998791536857325, 2.550254641757751, 2.8731038775629663, 2.865932456865462, 2.074754886007174, 2.163413033589733, 2.0387235923688944, 2.040521298664257], \"Total\": [75.0, 15.0, 39.0, 25.0, 20.0, 10.0, 12.0, 21.0, 10.0, 25.0, 14.0, 11.0, 9.0, 18.0, 17.0, 29.0, 10.0, 25.0, 6.0, 33.0, 18.0, 13.0, 12.0, 20.0, 86.0, 23.0, 8.0, 58.0, 21.0, 24.0, 21.19039335681883, 15.759691363698087, 9.64109794650696, 19.916264323214335, 12.176000229038943, 41.75048080673306, 8.752345934038795, 12.120969231297474, 7.841671698124239, 7.865035670021385, 6.9283572783209895, 6.918256957321183, 7.790277769463213, 6.982767806124686, 7.829790724461245, 18.858690096860368, 30.28014199331076, 6.025896265753955, 6.088601899501491, 5.989801871930119, 6.912465030153837, 9.310653803246122, 8.333686004344237, 5.990771320193872, 6.028240486524594, 6.1047676948434475, 6.892790284055939, 6.008210486943833, 6.893464671327681, 6.031986373166191, 20.43603825703463, 18.117989537912322, 11.816291429801058, 9.541314806861891, 86.62508381567112, 83.83869930620585, 58.04174079550064, 11.939512403567873, 16.10580303668648, 16.18587097996574, 58.861311574472694, 30.870681859879, 67.98078663678605, 36.233446371857035, 43.03022979965951, 19.55305557967275, 20.66411174172808, 27.031357434047536, 67.42534088314753, 30.556503517632557, 29.9206658266658, 73.80287260747964, 20.706082369432437, 34.80240065819228, 24.23793624958606, 37.73923501830619, 26.295428979234146, 75.24437623596049, 25.00624203147209, 39.4149169034317, 33.897303928387764, 27.745665874541483, 26.20140804200492, 6.380110248131464, 7.735207958312912, 7.13436131109542, 4.794578654707524, 4.7756879986889835, 7.23973922451736, 10.526702106439169, 3.955247606405421, 3.9957797562295543, 3.96567079340759, 5.294262513323695, 5.595005387622658, 6.489704160579293, 5.430513180314363, 9.049903678223256, 8.180457226330835, 3.171588085282374, 3.1489421514771823, 3.100860884296927, 3.1783070092992602, 3.209451031746459, 9.35816372025707, 3.1670541558570013, 3.1120715879889818, 3.142997581966767, 4.5800246299306, 5.5887565734700395, 3.2461989975173977, 3.189274021973363, 4.507685909849824, 8.808057009340589, 23.95653243337176, 21.10346088049872, 7.86970392282037, 10.160378937891338, 4.8392445864495315, 6.333757012314388, 75.24437623596049, 27.745665874541483, 13.61136409034526, 73.80287260747964, 14.03742917962428, 25.139611244399255, 67.42534088314753, 19.11606433255341, 86.62508381567112, 30.870681859879, 16.558314460734422, 67.98078663678605, 58.04174079550064, 83.83869930620585, 26.20140804200492, 21.92103473104775, 21.894996444259185, 37.73923501830619, 24.903859254998476, 27.511102325895074, 58.861311574472694, 36.233446371857035, 24.147655151774615, 21.964851771463593, 23.628289856622537, 25.357545866491115, 10.44967538571607, 4.2549349042369355, 5.566127879840937, 5.809180934290009, 4.778791129661143, 12.399744830089363, 6.055285354538593, 5.071149661570979, 2.902470214621675, 2.9559654021556914, 4.156263625375272, 8.063911057635021, 7.851568443197332, 4.348114897275911, 2.8917707018703616, 2.9205751437531933, 4.4101603668385465, 10.738309690699092, 2.9294490940560074, 2.9453837172653485, 6.471011670911985, 3.5100686217346078, 4.551506028412434, 3.6715993704263026, 11.19737167172186, 3.792647428153838, 5.20144860907833, 4.4331787496094295, 3.6807382109752846, 5.697104199203252, 14.03742917962428, 17.06027101562016, 11.941316307682076, 21.852123204644787, 5.330059658552321, 5.942527790903968, 14.062822952450109, 20.722537715249352, 8.190322991636664, 33.897303928387764, 27.511102325895074, 23.628289856622537, 24.903859254998476, 67.98078663678605, 9.469655125495711, 18.105166323606767, 58.861311574472694, 18.79600160538337, 73.80287260747964, 12.657045940627201, 19.36699167087921, 67.42534088314753, 39.4149169034317, 34.80240065819228, 83.83869930620585, 21.894996444259185, 25.357545866491115, 15.460332567280869, 3.521911052512598, 3.517204626150016, 9.174653178747935, 2.8574377391822403, 2.928513770204353, 2.7877537307923097, 2.8991355814492636, 2.889438772080766, 4.368448802946259, 8.022498825576758, 4.359250771030312, 6.0497588945832765, 3.3949044693239943, 3.5200949565468758, 5.095212688877851, 3.743914484507778, 3.7412001060662377, 3.7624731700175946, 3.737673230792439, 3.730803590445227, 3.7389167895261743, 7.9037997877772215, 7.998510624727997, 4.416559485968632, 8.903108348680757, 4.629300846380429, 2.7244320125685286, 4.60725574436641, 2.771343653355688, 11.19737167172186, 20.620989183257947, 18.045161946446285, 9.244512282137931, 39.4149169034317, 25.374958741304003, 15.116744923573233, 75.24437623596049, 9.672451629098319, 15.159348152135081, 24.147655151774615, 43.03022979965951, 12.817788281567001, 86.62508381567112, 27.031357434047536, 20.751639118397797, 83.83869930620585, 67.42534088314753, 37.73923501830619, 73.80287260747964, 36.233446371857035, 67.98078663678605, 34.80240065819228, 58.04174079550064, 58.861311574472694, 21.964851771463593, 33.897303928387764, 4.955146388942963, 6.13682887102165, 3.2468491275628324, 4.599475315405513, 2.574158281024186, 2.63740008195255, 2.642673321879144, 2.6839454494272923, 2.605165989675964, 2.6658213220916283, 2.702420310474383, 2.68017098818283, 2.7240991322933414, 4.736575444954372, 3.920963698545169, 2.7564592220262716, 10.088101444528007, 3.27199276606993, 3.3357751103684077, 3.446649322587767, 4.697952494457543, 3.5674429972610517, 3.500386930870997, 7.008054878504831, 7.122483552092323, 4.176589979874658, 7.573567758210336, 4.3511969996217275, 5.852676230470542, 2.7244320125685286, 4.033366240178784, 6.778291904352798, 13.174333425106656, 25.374958741304003, 75.24437623596049, 25.357545866491115, 12.401585256706, 8.951561834046505, 29.9206658266658, 7.647197245831091, 25.139611244399255, 14.45262136908535, 18.79600160538337, 58.861311574472694, 17.751884830965352, 8.087292501935366, 83.83869930620585, 39.4149169034317, 58.04174079550064, 34.80240065819228, 67.98078663678605, 86.62508381567112, 30.28014199331076, 73.80287260747964, 26.295428979234146, 30.556503517632557], \"logprob\": [30.0, 29.0, 28.0, 27.0, 26.0, 25.0, 24.0, 23.0, 22.0, 21.0, 20.0, 19.0, 18.0, 17.0, 16.0, 15.0, 14.0, 13.0, 12.0, 11.0, 10.0, 9.0, 8.0, 7.0, 6.0, 5.0, 4.0, 3.0, 2.0, 1.0, -5.0475, -5.3959, -5.8909, -5.1662, -5.664, -4.4364, -6.0069, -5.6947, -6.1353, -6.1377, -6.2702, -6.279, -6.1688, -6.2801, -6.167, -5.2898, -4.8238, -6.4444, -6.4351, -6.4534, -6.3131, -6.0183, -6.1293, -6.4599, -6.4574, -6.4455, -6.3244, -6.4623, -6.3253, -6.4642, -5.2551, -5.3801, -5.8004, -6.0174, -3.9308, -3.9813, -4.3466, -5.8136, -5.548, -5.5475, -4.3895, -4.9704, -4.3037, -4.8516, -4.7065, -5.3949, -5.3557, -5.1339, -4.3831, -5.0352, -5.0576, -4.3401, -5.3582, -4.9699, -5.2547, -4.9478, -5.2221, -4.5699, -5.2832, -5.0474, -5.1463, -5.2964, -5.3466, -5.5226, -5.3713, -5.4765, -5.8816, -5.9032, -5.5052, -5.1322, -6.1336, -6.1499, -6.1643, -5.8806, -5.8324, -5.6851, -5.8734, -5.3662, -5.5142, -6.4622, -6.4775, -6.4968, -6.4772, -6.469, -5.3993, -6.4862, -6.5073, -6.4995, -6.136, -5.9403, -6.484, -6.5108, -6.1919, -5.5478, -4.6386, -4.7786, -5.6876, -5.516, -6.1398, -5.9313, -3.9897, -4.7917, -5.341, -4.1197, -5.3668, -4.9586, -4.3478, -5.244, -4.3972, -5.0134, -5.3577, -4.6531, -4.7555, -4.6056, -5.1592, -5.241, -5.264, -5.064, -5.234, -5.248, -5.0859, -5.261, -5.3347, -5.3594, -5.3609, -5.3679, -4.4788, -5.5174, -5.3123, -5.3018, -5.5092, -4.5754, -5.3209, -5.5047, -6.0891, -6.0833, -5.7668, -5.1236, -5.1504, -5.7437, -6.1778, -6.1715, -5.7701, -4.8817, -6.2011, -6.2117, -5.4803, -6.1019, -5.8746, -6.1089, -5.0025, -6.0882, -5.7734, -5.9405, -6.1289, -5.7053, -4.8128, -4.6823, -5.1236, -4.7572, -5.8029, -5.7318, -5.1548, -4.9101, -5.523, -4.6084, -4.7827, -4.965, -4.9677, -4.462, -5.4794, -5.1928, -4.6963, -5.1908, -4.6587, -5.3606, -5.2267, -4.8115, -5.0523, -5.1111, -5.0871, -5.3052, -5.3067, -4.0999, -5.6396, -5.6499, -4.7394, -5.9867, -5.9711, -6.0231, -6.0186, -6.0308, -5.6229, -5.0201, -5.6557, -5.3724, -5.9773, -5.9957, -5.6368, -5.972, -5.9846, -5.9813, -5.9936, -6.0088, -6.0148, -5.3934, -5.3862, -5.9905, -5.327, -5.984, -6.521, -5.9965, -6.5206, -5.1473, -4.6121, -4.7659, -5.3804, -4.266, -4.687, -5.0873, -4.0693, -5.4127, -5.1452, -4.9025, -4.6546, -5.3565, -4.4894, -5.0409, -5.1694, -4.617, -4.7083, -4.9895, -4.8905, -5.1683, -5.1296, -5.2481, -5.2967, -5.313, -5.3618, -5.3573, -4.8991, -4.7422, -5.4951, -5.2464, -5.8506, -5.8397, -5.843, -5.831, -5.8641, -5.8486, -5.8383, -5.8466, -5.8406, -5.3269, -5.5384, -5.8989, -4.7072, -5.8576, -5.8528, -5.8419, -5.5337, -5.8263, -5.8815, -5.2453, -5.2338, -5.8093, -5.2175, -5.8251, -5.5733, -6.3531, -5.9756, -5.5322, -5.0472, -4.6367, -4.0372, -4.8165, -5.2015, -5.3902, -4.855, -5.4807, -4.9898, -5.2419, -5.1602, -4.8177, -5.2261, -5.483, -4.7609, -5.0201, -5.0319, -5.1939, -5.0747, -5.0772, -5.4003, -5.3584, -5.4178, -5.4169]}};\n",
       "\n",
       "function LDAvis_load_lib(url, callback){\n",
       "  var s = document.createElement('script');\n",
       "  s.src = url;\n",
       "  s.async = true;\n",
       "  s.onreadystatechange = s.onload = callback;\n",
       "  s.onerror = function(){console.warn(\"failed to load library \" + url);};\n",
       "  document.getElementsByTagName(\"head\")[0].appendChild(s);\n",
       "}\n",
       "\n",
       "if(typeof(LDAvis) !== \"undefined\"){\n",
       "   // already loaded: just create the visualization\n",
       "   !function(LDAvis){\n",
       "       new LDAvis(\"#\" + \"ldavis_el7836844925577765330763488\", ldavis_el7836844925577765330763488_data);\n",
       "   }(LDAvis);\n",
       "}else if(typeof define === \"function\" && define.amd){\n",
       "   // require.js is available: use it to load d3/LDAvis\n",
       "   require.config({paths: {d3: \"https://cdnjs.cloudflare.com/ajax/libs/d3/3.5.5/d3.min\"}});\n",
       "   require([\"d3\"], function(d3){\n",
       "      window.d3 = d3;\n",
       "      LDAvis_load_lib(\"https://cdn.rawgit.com/bmabey/pyLDAvis/files/ldavis.v1.0.0.js\", function(){\n",
       "        new LDAvis(\"#\" + \"ldavis_el7836844925577765330763488\", ldavis_el7836844925577765330763488_data);\n",
       "      });\n",
       "    });\n",
       "}else{\n",
       "    // require.js not available: dynamically load d3 & LDAvis\n",
       "    LDAvis_load_lib(\"https://cdnjs.cloudflare.com/ajax/libs/d3/3.5.5/d3.min.js\", function(){\n",
       "         LDAvis_load_lib(\"https://cdn.rawgit.com/bmabey/pyLDAvis/files/ldavis.v1.0.0.js\", function(){\n",
       "                 new LDAvis(\"#\" + \"ldavis_el7836844925577765330763488\", ldavis_el7836844925577765330763488_data);\n",
       "            })\n",
       "         });\n",
       "}\n",
       "</script>"
      ],
      "text/plain": [
       "PreparedData(topic_coordinates=            Freq  cluster  topics         x         y\n",
       "topic                                                \n",
       "4      49.931919        1       1  0.033673  0.014841\n",
       "2      20.782000        1       2  0.011932  0.071924\n",
       "0      11.774719        1       3 -0.081213 -0.012134\n",
       "3       9.973946        1       4  0.049986 -0.053080\n",
       "1       7.537415        1       5 -0.014378 -0.021551, topic_info=     Category       Freq        Term      Total  loglift  logprob\n",
       "term                                                             \n",
       "380   Default  75.000000       learn  75.000000  30.0000  30.0000\n",
       "366   Default  15.000000      kernel  15.000000  29.0000  29.0000\n",
       "290   Default  39.000000    function  39.000000  28.0000  28.0000\n",
       "207   Default  25.000000   distribut  25.000000  27.0000  27.0000\n",
       "104   Default  20.000000     classif  20.000000  26.0000  26.0000\n",
       "393   Default  10.000000         log  10.000000  25.0000  25.0000\n",
       "264   Default  12.000000      factor  12.000000  24.0000  24.0000\n",
       "410   Default  21.000000      markov  21.000000  23.0000  23.0000\n",
       "586   Default  10.000000     regress  10.000000  22.0000  22.0000\n",
       "296   Default  25.000000       gener  25.000000  21.0000  21.0000\n",
       "765   Default  14.000000      variat  14.000000  20.0000  20.0000\n",
       "515   Default  11.000000       point  11.000000  19.0000  19.0000\n",
       "160   Default   9.000000        cost   9.000000  18.0000  18.0000\n",
       "720   Default  18.000000     theoret  18.000000  17.0000  17.0000\n",
       "131   Default  17.000000     complex  17.000000  16.0000  16.0000\n",
       "487   Default  29.000000       optim  29.000000  15.0000  15.0000\n",
       "385   Default  10.000000  likelihood  10.000000  14.0000  14.0000\n",
       "80    Default  25.000000       bound  25.000000  13.0000  13.0000\n",
       "232   Default   6.000000     ensembl   6.000000  12.0000  12.0000\n",
       "623   Default  33.000000       sampl  33.000000  11.0000  11.0000\n",
       "729   Default  18.000000        time  18.000000  10.0000  10.0000\n",
       "548   Default  13.000000    properti  13.000000   9.0000   9.0000\n",
       "721   Default  12.000000      theori  12.000000   8.0000   8.0000\n",
       "388   Default  20.000000      linear  20.000000   7.0000   7.0000\n",
       "170   Default  86.000000        data  86.000000   6.0000   6.0000\n",
       "396   Default  23.000000         low  23.000000   5.0000   5.0000\n",
       "639   Default   8.000000      sensit   8.000000   4.0000   4.0000\n",
       "421   Default  58.000000      method  58.000000   3.0000   3.0000\n",
       "288   Default  21.000000   framework  21.000000   2.0000   2.0000\n",
       "463   Default  24.000000         new  24.000000   1.0000   1.0000\n",
       "...       ...        ...         ...        ...      ...      ...\n",
       "447    Topic5   2.490697         mrf   7.573568   1.4732  -5.2175\n",
       "601    Topic5   1.356569    research   4.351197   1.4198  -5.8251\n",
       "593    Topic5   1.745121       relax   5.852676   1.3752  -5.5733\n",
       "223    Topic5   0.800099   efficient   2.724432   1.3600  -6.3531\n",
       "97     Topic5   1.167101   character   4.033366   1.3452  -5.9756\n",
       "576    Topic5   1.818280    recoveri   6.778292   1.2695  -5.5322\n",
       "548    Topic5   2.953338    properti  13.174333   1.0900  -5.0472\n",
       "207    Topic5   4.452019   distribut  25.374959   0.8449  -4.6367\n",
       "380    Topic5   8.108164       learn  75.244376   0.3574  -4.0372\n",
       "296    Topic5   3.719617       gener  25.357546   0.6658  -4.8165\n",
       "721    Topic5   2.530863      theori  12.401585   0.9960  -5.2015\n",
       "382    Topic5   2.095699    learning   8.951562   1.1333  -5.3902\n",
       "487    Topic5   3.579213       optim  29.920666   0.4619  -4.8550\n",
       "575    Topic5   1.914365       recov   7.647197   1.2003  -5.4807\n",
       "80     Topic5   3.127735       bound  25.139611   0.5012  -4.9898\n",
       "335    Topic5   2.430742      improv  14.452621   0.8026  -5.2419\n",
       "729    Topic5   2.637602        time  18.796002   0.6215  -5.1602\n",
       "421    Topic5   3.714914      method  58.861312  -0.1775  -4.8177\n",
       "675    Topic5   2.469571       spars  17.751885   0.6128  -5.2261\n",
       "147    Topic5   1.910062  constraint   8.087293   1.1421  -5.4830\n",
       "724    Topic5   3.932112         thi  83.838699  -0.4744  -4.7609\n",
       "290    Topic5   3.034285    function  39.414917   0.0211  -5.0201\n",
       "535    Topic5   2.998792     problem  58.041741  -0.3777  -5.0319\n",
       "239    Topic5   2.550255       estim  34.802401  -0.0282  -5.1939\n",
       "440    Topic5   2.873104       model  67.980787  -0.5785  -5.0747\n",
       "170    Topic5   2.865932        data  86.625084  -0.8234  -5.0772\n",
       "399    Topic5   2.074755      machin  30.280142  -0.0954  -5.4003\n",
       "23     Topic5   2.163413   algorithm  73.802873  -0.9444  -5.3584\n",
       "607    Topic5   2.038724      result  26.295429   0.0282  -5.4178\n",
       "497    Topic5   2.040521       paper  30.556504  -0.1211  -5.4169\n",
       "\n",
       "[326 rows x 6 columns], token_table=      Topic      Freq       Term\n",
       "term                            \n",
       "22        1  0.834752    algebra\n",
       "23        1  0.541984  algorithm\n",
       "23        2  0.284542  algorithm\n",
       "23        3  0.094847  algorithm\n",
       "23        4  0.067748  algorithm\n",
       "23        5  0.027099  algorithm\n",
       "31        1  0.890015        ani\n",
       "32        1  0.829429    anomali\n",
       "33        1  0.497766      appli\n",
       "33        2  0.142219      appli\n",
       "33        3  0.284438      appli\n",
       "33        4  0.071109      appli\n",
       "36        1  0.582948   approach\n",
       "36        2  0.211981   approach\n",
       "36        3  0.052995   approach\n",
       "36        4  0.105990   approach\n",
       "36        5  0.052995   approach\n",
       "37        1  0.313872   approxim\n",
       "37        2  0.366184   approxim\n",
       "37        3  0.156936   approxim\n",
       "37        4  0.104624   approxim\n",
       "37        5  0.052312   approxim\n",
       "43        1  0.154090        arm\n",
       "43        2  0.616361        arm\n",
       "51        1  0.280801   asymptot\n",
       "51        2  0.280801   asymptot\n",
       "51        5  0.280801   asymptot\n",
       "52        4  0.682940   attribut\n",
       "58        1  0.189993     bandit\n",
       "58        2  0.664976     bandit\n",
       "...     ...       ...        ...\n",
       "736       1  0.677503      train\n",
       "736       2  0.096786      train\n",
       "736       3  0.096786      train\n",
       "736       4  0.096786      train\n",
       "738       1  0.235021    transit\n",
       "738       3  0.705064    transit\n",
       "740       1  0.725396       tree\n",
       "740       4  0.145079       tree\n",
       "749       1  0.217416    uniform\n",
       "749       5  0.434832    uniform\n",
       "755       1  0.563586        use\n",
       "755       2  0.237300        use\n",
       "755       3  0.088987        use\n",
       "755       4  0.074156        use\n",
       "755       5  0.029662        use\n",
       "765       1  0.142476     variat\n",
       "765       2  0.427429     variat\n",
       "765       3  0.427429     variat\n",
       "769       3  0.689068    vectors\n",
       "771       1  0.375230    version\n",
       "771       3  0.375230    version\n",
       "774       1  0.766304      video\n",
       "780       2  0.736579       weak\n",
       "780       4  0.184145       weak\n",
       "781       3  0.284083        web\n",
       "781       4  0.568166        web\n",
       "784       1  0.439415       wide\n",
       "784       3  0.439415       wide\n",
       "785       1  0.265783       word\n",
       "785       4  0.531565       word\n",
       "\n",
       "[621 rows x 3 columns], R=30, lambda_step=0.01, plot_opts={'xlab': 'PC1', 'ylab': 'PC2'}, topic_order=[5, 3, 1, 4, 2])"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/Cellar/python/2.7.13/Frameworks/Python.framework/Versions/2.7/lib/python2.7/json/encoder.py:207: DeprecationWarning: Interpreting naive datetime as local 2017-04-02 18:07:49.170864. Please add timezone info to timestamps.\n",
      "  chunks = self.iterencode(o, _one_shot=True)\n"
     ]
    }
   ],
   "source": [
    "pyLDAvis.sklearn.prepare(lda, corpus_counter, counter)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "Reference:\n",
    "\n",
    "[1] Deerwester, S., Dumais, S. T., Furnas, G. W., Landauer, T. K., & Harshman, R. (1990). Indexing by latent semantic analysis. Journal of the American society for information science, 41(6), 391.\n",
    "\n",
    "[2] Blei, D. M., Ng, A. Y., & Jordan, M. I. (2003). Latent dirichlet allocation. Journal of machine Learning research, 3(Jan), 993-1022."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}