7-applying/3-similarity.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Part 7.3 - Document Similarity on Topic Modeling\n",
    "---\n",
    "\n",
    "### Papers Past Topic Modeling\n",
    "<br/>\n",
    "\n",
    "Ben Faulks - bmf43@uclive.ac.nz\n",
    "\n",
    "Xiandong Cai - xca24@uclive.ac.nz\n",
    "\n",
    "Yujie Cui - ycu23@uclive.ac.nz"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2019-02-03 20:24:37\n",
      "[('spark.app.id', 'local-1549178690275'),\n",
      " ('spark.app.name', 'local'),\n",
      " ('spark.driver.maxResultSize', '8g'),\n",
      " ('spark.rdd.compress', 'True'),\n",
      " ('spark.driver.port', '42474'),\n",
      " ('spark.driver.host', '192.168.1.207'),\n",
      " ('spark.driver.memory', '62g'),\n",
      " ('spark.master', 'local[*]'),\n",
      " ('spark.executor.id', 'driver'),\n",
      " ('spark.submit.deployMode', 'client'),\n",
      " ('spark.serializer.objectStreamReset', '100'),\n",
      " ('spark.ui.showConsoleProgress', 'true'),\n",
      " ('spark.driver.cores', '6')]\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "\n",
       "        <div>\n",
       "            <p><b>SparkContext</b></p>\n",
       "\n",
       "            <p><a href=\"http://192.168.1.207:4040\">Spark UI</a></p>\n",
       "\n",
       "            <dl>\n",
       "              <dt>Version</dt>\n",
       "                <dd><code>v2.4.0</code></dd>\n",
       "              <dt>Master</dt>\n",
       "                <dd><code>local[*]</code></dd>\n",
       "              <dt>AppName</dt>\n",
       "                <dd><code>local</code></dd>\n",
       "            </dl>\n",
       "        </div>\n",
       "        "
      ],
      "text/plain": [
       "<SparkContext master=local[*] appName=local>"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import sys, gc\n",
    "sys.path.insert(0, '../utils')\n",
    "from utils import conf_pyspark, load_dataset\n",
    "from utils_preplot import load_doctopic\n",
    "#import warnings\n",
    "#warnings.filterwarnings('ignore')\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from scipy.stats import entropy\n",
    "pd.set_option('display.max_columns', 120)\n",
    "pd.set_option('display.max_colwidth', 120)\n",
    "from matplotlib import pyplot as plt\n",
    "plt.style.use('ggplot')\n",
    "%matplotlib inline\n",
    "dpi=80\n",
    "\n",
    "import datetime\n",
    "print (datetime.datetime.now().strftime(\"%Y-%m-%d %H:%M:%S\"))\n",
    "\n",
    "from pyspark.sql import functions as F\n",
    "from pyspark.sql.types import *\n",
    "\n",
    "# intiate PySpark\n",
    "sc, spark = conf_pyspark()\n",
    "\n",
    "sc"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**You read an article in the corpus, what if you want to know more about similar articles? In this part, we will calculate the similarity based on topic modeling and recommand similar articles for you.**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1 Load Data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Meta Dataset:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>publisher</th>\n",
       "      <th>region</th>\n",
       "      <th>date</th>\n",
       "      <th>ads</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1854213</td>\n",
       "      <td>Albertland Gazette</td>\n",
       "      <td>Auckland</td>\n",
       "      <td>1862-06-14</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1854215</td>\n",
       "      <td>Albertland Gazette</td>\n",
       "      <td>Auckland</td>\n",
       "      <td>1862-06-14</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1854221</td>\n",
       "      <td>Albertland Gazette</td>\n",
       "      <td>Auckland</td>\n",
       "      <td>1862-06-14</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1854224</td>\n",
       "      <td>Albertland Gazette</td>\n",
       "      <td>Auckland</td>\n",
       "      <td>1862-07-03</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1854232</td>\n",
       "      <td>Albertland Gazette</td>\n",
       "      <td>Auckland</td>\n",
       "      <td>1863-08-01</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        id           publisher    region        date    ads\n",
       "0  1854213  Albertland Gazette  Auckland  1862-06-14  False\n",
       "1  1854215  Albertland Gazette  Auckland  1862-06-14   True\n",
       "2  1854221  Albertland Gazette  Auckland  1862-06-14  False\n",
       "3  1854224  Albertland Gazette  Auckland  1862-07-03  False\n",
       "4  1854232  Albertland Gazette  Auckland  1863-08-01   True"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_meta = load_dataset('meta', spark)\n",
    "df_meta = df_meta.toPandas()\n",
    "df_meta.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Topic List:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape of dataframe: (200, 3)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>topic_</th>\n",
       "      <th>weight_</th>\n",
       "      <th>keywords_</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>82</th>\n",
       "      <td>82</td>\n",
       "      <td>0.00688</td>\n",
       "      <td>sale apply land acres house particulars property terms town good sections situated lease section years farm road fre...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>93</th>\n",
       "      <td>93</td>\n",
       "      <td>0.00339</td>\n",
       "      <td>good race mile furlongs ran won day straight run home half horse handicap field horses back front winner time win</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101</th>\n",
       "      <td>101</td>\n",
       "      <td>0.01135</td>\n",
       "      <td>years london great man english time england made ago year life lady young world day french long paris found days</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>54</th>\n",
       "      <td>54</td>\n",
       "      <td>0.00551</td>\n",
       "      <td>church bishop catholic rev father synod religious roman archbishop clergy catholics pope parish cathedral england di...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>108</th>\n",
       "      <td>108</td>\n",
       "      <td>0.00765</td>\n",
       "      <td>auction sale o'clock sell public instructions day saturday rooms auctioneers auctioneer wednesday received messrs lo...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     topic_  weight_  \\\n",
       "82       82  0.00688   \n",
       "93       93  0.00339   \n",
       "101     101  0.01135   \n",
       "54       54  0.00551   \n",
       "108     108  0.00765   \n",
       "\n",
       "                                                                                                                   keywords_  \n",
       "82   sale apply land acres house particulars property terms town good sections situated lease section years farm road fre...  \n",
       "93        good race mile furlongs ran won day straight run home half horse handicap field horses back front winner time win   \n",
       "101        years london great man english time england made ago year life lady young world day french long paris found days   \n",
       "54   church bishop catholic rev father synod religious roman archbishop clergy catholics pope parish cathedral england di...  \n",
       "108  auction sale o'clock sell public instructions day saturday rooms auctioneers auctioneer wednesday received messrs lo...  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "path = r'../models/train/topicKeys.txt'\n",
    "\n",
    "df_topics = (pd.read_table(path, \n",
    "                           header=None, \n",
    "                           index_col=None, \n",
    "                           names = ['topic_', 'weight_', 'keywords_'],\n",
    "                           encoding='utf8')\n",
    "             .sort_values(by='topic_')\n",
    "             .reset_index(drop=True))\n",
    "\n",
    "topics_shape = df_topics.shape\n",
    "topic_number = df_topics.shape[0]\n",
    "\n",
    "print('Shape of dataframe:', topics_shape)\n",
    "df_topics.sample(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Load doc-topic matrix:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "path = r'../models/train/docTopics.txt'\n",
    "\n",
    "df_doctopic = (load_doctopic(path, topics_shape[0], spark)).drop('index').orderBy('id').drop('id').toPandas()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape of dataframe: (3025602, 200)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>10</th>\n",
       "      <th>11</th>\n",
       "      <th>12</th>\n",
       "      <th>13</th>\n",
       "      <th>14</th>\n",
       "      <th>15</th>\n",
       "      <th>16</th>\n",
       "      <th>17</th>\n",
       "      <th>18</th>\n",
       "      <th>19</th>\n",
       "      <th>20</th>\n",
       "      <th>21</th>\n",
       "      <th>22</th>\n",
       "      <th>23</th>\n",
       "      <th>24</th>\n",
       "      <th>25</th>\n",
       "      <th>26</th>\n",
       "      <th>27</th>\n",
       "      <th>28</th>\n",
       "      <th>29</th>\n",
       "      <th>30</th>\n",
       "      <th>31</th>\n",
       "      <th>32</th>\n",
       "      <th>33</th>\n",
       "      <th>34</th>\n",
       "      <th>35</th>\n",
       "      <th>36</th>\n",
       "      <th>37</th>\n",
       "      <th>38</th>\n",
       "      <th>39</th>\n",
       "      <th>40</th>\n",
       "      <th>41</th>\n",
       "      <th>42</th>\n",
       "      <th>43</th>\n",
       "      <th>44</th>\n",
       "      <th>45</th>\n",
       "      <th>46</th>\n",
       "      <th>47</th>\n",
       "      <th>48</th>\n",
       "      <th>49</th>\n",
       "      <th>50</th>\n",
       "      <th>51</th>\n",
       "      <th>52</th>\n",
       "      <th>53</th>\n",
       "      <th>54</th>\n",
       "      <th>55</th>\n",
       "      <th>56</th>\n",
       "      <th>57</th>\n",
       "      <th>58</th>\n",
       "      <th>59</th>\n",
       "      <th>...</th>\n",
       "      <th>140</th>\n",
       "      <th>141</th>\n",
       "      <th>142</th>\n",
       "      <th>143</th>\n",
       "      <th>144</th>\n",
       "      <th>145</th>\n",
       "      <th>146</th>\n",
       "      <th>147</th>\n",
       "      <th>148</th>\n",
       "      <th>149</th>\n",
       "      <th>150</th>\n",
       "      <th>151</th>\n",
       "      <th>152</th>\n",
       "      <th>153</th>\n",
       "      <th>154</th>\n",
       "      <th>155</th>\n",
       "      <th>156</th>\n",
       "      <th>157</th>\n",
       "      <th>158</th>\n",
       "      <th>159</th>\n",
       "      <th>160</th>\n",
       "      <th>161</th>\n",
       "      <th>162</th>\n",
       "      <th>163</th>\n",
       "      <th>164</th>\n",
       "      <th>165</th>\n",
       "      <th>166</th>\n",
       "      <th>167</th>\n",
       "      <th>168</th>\n",
       "      <th>169</th>\n",
       "      <th>170</th>\n",
       "      <th>171</th>\n",
       "      <th>172</th>\n",
       "      <th>173</th>\n",
       "      <th>174</th>\n",
       "      <th>175</th>\n",
       "      <th>176</th>\n",
       "      <th>177</th>\n",
       "      <th>178</th>\n",
       "      <th>179</th>\n",
       "      <th>180</th>\n",
       "      <th>181</th>\n",
       "      <th>182</th>\n",
       "      <th>183</th>\n",
       "      <th>184</th>\n",
       "      <th>185</th>\n",
       "      <th>186</th>\n",
       "      <th>187</th>\n",
       "      <th>188</th>\n",
       "      <th>189</th>\n",
       "      <th>190</th>\n",
       "      <th>191</th>\n",
       "      <th>192</th>\n",
       "      <th>193</th>\n",
       "      <th>194</th>\n",
       "      <th>195</th>\n",
       "      <th>196</th>\n",
       "      <th>197</th>\n",
       "      <th>198</th>\n",
       "      <th>199</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.000196</td>\n",
       "      <td>0.000321</td>\n",
       "      <td>0.777824</td>\n",
       "      <td>0.000115</td>\n",
       "      <td>0.000269</td>\n",
       "      <td>0.000154</td>\n",
       "      <td>0.000291</td>\n",
       "      <td>0.000227</td>\n",
       "      <td>0.000227</td>\n",
       "      <td>0.000143</td>\n",
       "      <td>0.000254</td>\n",
       "      <td>0.000210</td>\n",
       "      <td>0.000139</td>\n",
       "      <td>0.000235</td>\n",
       "      <td>0.000221</td>\n",
       "      <td>0.000279</td>\n",
       "      <td>0.000245</td>\n",
       "      <td>0.000247</td>\n",
       "      <td>0.000178</td>\n",
       "      <td>0.000305</td>\n",
       "      <td>0.000135</td>\n",
       "      <td>0.090193</td>\n",
       "      <td>0.000215</td>\n",
       "      <td>0.000042</td>\n",
       "      <td>0.000103</td>\n",
       "      <td>0.000260</td>\n",
       "      <td>0.000069</td>\n",
       "      <td>0.000106</td>\n",
       "      <td>0.000294</td>\n",
       "      <td>0.000197</td>\n",
       "      <td>0.000227</td>\n",
       "      <td>0.000176</td>\n",
       "      <td>0.000121</td>\n",
       "      <td>0.031044</td>\n",
       "      <td>0.000048</td>\n",
       "      <td>0.000190</td>\n",
       "      <td>0.000268</td>\n",
       "      <td>0.000177</td>\n",
       "      <td>0.000158</td>\n",
       "      <td>0.000230</td>\n",
       "      <td>0.000169</td>\n",
       "      <td>0.000165</td>\n",
       "      <td>0.000215</td>\n",
       "      <td>0.000315</td>\n",
       "      <td>0.000183</td>\n",
       "      <td>0.000184</td>\n",
       "      <td>0.000214</td>\n",
       "      <td>0.000159</td>\n",
       "      <td>0.000065</td>\n",
       "      <td>0.000088</td>\n",
       "      <td>0.000263</td>\n",
       "      <td>0.000187</td>\n",
       "      <td>0.000139</td>\n",
       "      <td>0.000104</td>\n",
       "      <td>0.000165</td>\n",
       "      <td>0.000172</td>\n",
       "      <td>0.000155</td>\n",
       "      <td>0.000288</td>\n",
       "      <td>0.000312</td>\n",
       "      <td>0.000243</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000170</td>\n",
       "      <td>0.000325</td>\n",
       "      <td>0.000338</td>\n",
       "      <td>0.000191</td>\n",
       "      <td>0.000035</td>\n",
       "      <td>0.000223</td>\n",
       "      <td>0.000237</td>\n",
       "      <td>0.000350</td>\n",
       "      <td>0.000084</td>\n",
       "      <td>0.000255</td>\n",
       "      <td>0.000391</td>\n",
       "      <td>0.000196</td>\n",
       "      <td>0.000191</td>\n",
       "      <td>0.000165</td>\n",
       "      <td>0.000213</td>\n",
       "      <td>0.000213</td>\n",
       "      <td>0.000292</td>\n",
       "      <td>0.000288</td>\n",
       "      <td>0.000261</td>\n",
       "      <td>0.000406</td>\n",
       "      <td>0.000239</td>\n",
       "      <td>0.000134</td>\n",
       "      <td>0.000109</td>\n",
       "      <td>0.000175</td>\n",
       "      <td>0.000244</td>\n",
       "      <td>0.000219</td>\n",
       "      <td>0.000119</td>\n",
       "      <td>0.000173</td>\n",
       "      <td>0.000272</td>\n",
       "      <td>0.000128</td>\n",
       "      <td>0.000257</td>\n",
       "      <td>0.000112</td>\n",
       "      <td>0.000072</td>\n",
       "      <td>0.000100</td>\n",
       "      <td>0.000155</td>\n",
       "      <td>0.000189</td>\n",
       "      <td>0.000134</td>\n",
       "      <td>0.000111</td>\n",
       "      <td>0.000168</td>\n",
       "      <td>0.000158</td>\n",
       "      <td>0.000222</td>\n",
       "      <td>0.000211</td>\n",
       "      <td>0.000307</td>\n",
       "      <td>0.000206</td>\n",
       "      <td>0.000256</td>\n",
       "      <td>0.000198</td>\n",
       "      <td>0.000159</td>\n",
       "      <td>0.000438</td>\n",
       "      <td>0.000215</td>\n",
       "      <td>0.000215</td>\n",
       "      <td>0.000269</td>\n",
       "      <td>0.000322</td>\n",
       "      <td>0.000177</td>\n",
       "      <td>0.000282</td>\n",
       "      <td>0.000290</td>\n",
       "      <td>0.000466</td>\n",
       "      <td>0.000188</td>\n",
       "      <td>0.000201</td>\n",
       "      <td>0.000230</td>\n",
       "      <td>0.000148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.000112</td>\n",
       "      <td>0.000184</td>\n",
       "      <td>0.000379</td>\n",
       "      <td>0.000066</td>\n",
       "      <td>0.000154</td>\n",
       "      <td>0.000088</td>\n",
       "      <td>0.000166</td>\n",
       "      <td>0.000130</td>\n",
       "      <td>0.102773</td>\n",
       "      <td>0.000082</td>\n",
       "      <td>0.000146</td>\n",
       "      <td>0.000120</td>\n",
       "      <td>0.000080</td>\n",
       "      <td>0.000135</td>\n",
       "      <td>0.000126</td>\n",
       "      <td>0.000160</td>\n",
       "      <td>0.000140</td>\n",
       "      <td>0.000141</td>\n",
       "      <td>0.000102</td>\n",
       "      <td>0.000174</td>\n",
       "      <td>0.000077</td>\n",
       "      <td>0.000298</td>\n",
       "      <td>0.000123</td>\n",
       "      <td>0.000024</td>\n",
       "      <td>0.000059</td>\n",
       "      <td>0.000149</td>\n",
       "      <td>0.000040</td>\n",
       "      <td>0.000061</td>\n",
       "      <td>0.000168</td>\n",
       "      <td>0.000113</td>\n",
       "      <td>0.000130</td>\n",
       "      <td>0.000101</td>\n",
       "      <td>0.000069</td>\n",
       "      <td>0.034874</td>\n",
       "      <td>0.000028</td>\n",
       "      <td>0.000109</td>\n",
       "      <td>0.000153</td>\n",
       "      <td>0.000101</td>\n",
       "      <td>0.000091</td>\n",
       "      <td>0.000131</td>\n",
       "      <td>0.000097</td>\n",
       "      <td>0.000095</td>\n",
       "      <td>0.000123</td>\n",
       "      <td>0.000180</td>\n",
       "      <td>0.000104</td>\n",
       "      <td>0.000106</td>\n",
       "      <td>0.000123</td>\n",
       "      <td>0.000091</td>\n",
       "      <td>0.000037</td>\n",
       "      <td>0.000050</td>\n",
       "      <td>0.000151</td>\n",
       "      <td>0.000107</td>\n",
       "      <td>0.000080</td>\n",
       "      <td>0.000060</td>\n",
       "      <td>0.000094</td>\n",
       "      <td>0.000098</td>\n",
       "      <td>0.000089</td>\n",
       "      <td>0.000165</td>\n",
       "      <td>0.000178</td>\n",
       "      <td>0.000139</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000097</td>\n",
       "      <td>0.000186</td>\n",
       "      <td>0.000193</td>\n",
       "      <td>0.000109</td>\n",
       "      <td>0.000020</td>\n",
       "      <td>0.000128</td>\n",
       "      <td>0.000135</td>\n",
       "      <td>0.000200</td>\n",
       "      <td>0.000048</td>\n",
       "      <td>0.000146</td>\n",
       "      <td>0.000224</td>\n",
       "      <td>0.000112</td>\n",
       "      <td>0.000109</td>\n",
       "      <td>0.000094</td>\n",
       "      <td>0.000122</td>\n",
       "      <td>0.325158</td>\n",
       "      <td>0.000167</td>\n",
       "      <td>0.000165</td>\n",
       "      <td>0.000149</td>\n",
       "      <td>0.000232</td>\n",
       "      <td>0.000137</td>\n",
       "      <td>0.000077</td>\n",
       "      <td>0.000062</td>\n",
       "      <td>0.000100</td>\n",
       "      <td>0.000140</td>\n",
       "      <td>0.000125</td>\n",
       "      <td>0.000068</td>\n",
       "      <td>0.000099</td>\n",
       "      <td>0.000156</td>\n",
       "      <td>0.000073</td>\n",
       "      <td>0.000147</td>\n",
       "      <td>0.000064</td>\n",
       "      <td>0.000041</td>\n",
       "      <td>0.000057</td>\n",
       "      <td>0.000089</td>\n",
       "      <td>0.000108</td>\n",
       "      <td>0.000077</td>\n",
       "      <td>0.000064</td>\n",
       "      <td>0.000096</td>\n",
       "      <td>0.000090</td>\n",
       "      <td>0.000127</td>\n",
       "      <td>0.000121</td>\n",
       "      <td>0.000176</td>\n",
       "      <td>0.000118</td>\n",
       "      <td>0.000147</td>\n",
       "      <td>0.000113</td>\n",
       "      <td>0.000091</td>\n",
       "      <td>0.000251</td>\n",
       "      <td>0.000123</td>\n",
       "      <td>0.000123</td>\n",
       "      <td>0.000154</td>\n",
       "      <td>0.000184</td>\n",
       "      <td>0.051423</td>\n",
       "      <td>0.000161</td>\n",
       "      <td>0.000166</td>\n",
       "      <td>0.000267</td>\n",
       "      <td>0.000108</td>\n",
       "      <td>0.000115</td>\n",
       "      <td>0.000131</td>\n",
       "      <td>0.000085</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.000158</td>\n",
       "      <td>0.000259</td>\n",
       "      <td>0.000534</td>\n",
       "      <td>0.000093</td>\n",
       "      <td>0.000217</td>\n",
       "      <td>0.000124</td>\n",
       "      <td>0.000234</td>\n",
       "      <td>0.000184</td>\n",
       "      <td>0.000183</td>\n",
       "      <td>0.000116</td>\n",
       "      <td>0.000205</td>\n",
       "      <td>0.000169</td>\n",
       "      <td>0.000112</td>\n",
       "      <td>0.000190</td>\n",
       "      <td>0.000178</td>\n",
       "      <td>0.000225</td>\n",
       "      <td>0.000198</td>\n",
       "      <td>0.000199</td>\n",
       "      <td>0.000144</td>\n",
       "      <td>0.000246</td>\n",
       "      <td>0.000109</td>\n",
       "      <td>0.000420</td>\n",
       "      <td>0.000173</td>\n",
       "      <td>0.000034</td>\n",
       "      <td>0.000083</td>\n",
       "      <td>0.000210</td>\n",
       "      <td>0.000056</td>\n",
       "      <td>0.000086</td>\n",
       "      <td>0.000237</td>\n",
       "      <td>0.000159</td>\n",
       "      <td>0.000183</td>\n",
       "      <td>0.000142</td>\n",
       "      <td>0.000098</td>\n",
       "      <td>0.025053</td>\n",
       "      <td>0.000039</td>\n",
       "      <td>0.000154</td>\n",
       "      <td>0.000216</td>\n",
       "      <td>0.892676</td>\n",
       "      <td>0.000128</td>\n",
       "      <td>0.000185</td>\n",
       "      <td>0.000137</td>\n",
       "      <td>0.000133</td>\n",
       "      <td>0.000173</td>\n",
       "      <td>0.000254</td>\n",
       "      <td>0.000147</td>\n",
       "      <td>0.000149</td>\n",
       "      <td>0.000173</td>\n",
       "      <td>0.000128</td>\n",
       "      <td>0.000052</td>\n",
       "      <td>0.000071</td>\n",
       "      <td>0.000213</td>\n",
       "      <td>0.000151</td>\n",
       "      <td>0.000112</td>\n",
       "      <td>0.000084</td>\n",
       "      <td>0.000133</td>\n",
       "      <td>0.000139</td>\n",
       "      <td>0.000125</td>\n",
       "      <td>0.000233</td>\n",
       "      <td>0.000252</td>\n",
       "      <td>0.000196</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000137</td>\n",
       "      <td>0.000263</td>\n",
       "      <td>0.000273</td>\n",
       "      <td>0.000154</td>\n",
       "      <td>0.000028</td>\n",
       "      <td>0.000180</td>\n",
       "      <td>0.000191</td>\n",
       "      <td>0.000282</td>\n",
       "      <td>0.000068</td>\n",
       "      <td>0.000206</td>\n",
       "      <td>0.000316</td>\n",
       "      <td>0.000158</td>\n",
       "      <td>0.000154</td>\n",
       "      <td>0.000133</td>\n",
       "      <td>0.000172</td>\n",
       "      <td>0.000172</td>\n",
       "      <td>0.000236</td>\n",
       "      <td>0.000232</td>\n",
       "      <td>0.000210</td>\n",
       "      <td>0.024450</td>\n",
       "      <td>0.000193</td>\n",
       "      <td>0.000108</td>\n",
       "      <td>0.000088</td>\n",
       "      <td>0.000141</td>\n",
       "      <td>0.000197</td>\n",
       "      <td>0.000177</td>\n",
       "      <td>0.000096</td>\n",
       "      <td>0.000139</td>\n",
       "      <td>0.000220</td>\n",
       "      <td>0.000103</td>\n",
       "      <td>0.000208</td>\n",
       "      <td>0.000091</td>\n",
       "      <td>0.000058</td>\n",
       "      <td>0.000081</td>\n",
       "      <td>0.000125</td>\n",
       "      <td>0.000153</td>\n",
       "      <td>0.000108</td>\n",
       "      <td>0.000090</td>\n",
       "      <td>0.000136</td>\n",
       "      <td>0.000127</td>\n",
       "      <td>0.000179</td>\n",
       "      <td>0.000170</td>\n",
       "      <td>0.000248</td>\n",
       "      <td>0.000166</td>\n",
       "      <td>0.000207</td>\n",
       "      <td>0.000159</td>\n",
       "      <td>0.000129</td>\n",
       "      <td>0.000354</td>\n",
       "      <td>0.000174</td>\n",
       "      <td>0.000174</td>\n",
       "      <td>0.000217</td>\n",
       "      <td>0.000260</td>\n",
       "      <td>0.000143</td>\n",
       "      <td>0.000228</td>\n",
       "      <td>0.000234</td>\n",
       "      <td>0.000376</td>\n",
       "      <td>0.000152</td>\n",
       "      <td>0.000162</td>\n",
       "      <td>0.000185</td>\n",
       "      <td>0.000120</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.000148</td>\n",
       "      <td>0.000242</td>\n",
       "      <td>0.000498</td>\n",
       "      <td>0.000086</td>\n",
       "      <td>0.000203</td>\n",
       "      <td>0.000116</td>\n",
       "      <td>0.000219</td>\n",
       "      <td>0.000171</td>\n",
       "      <td>0.000171</td>\n",
       "      <td>0.000108</td>\n",
       "      <td>0.000191</td>\n",
       "      <td>0.000158</td>\n",
       "      <td>0.000105</td>\n",
       "      <td>0.000177</td>\n",
       "      <td>0.000166</td>\n",
       "      <td>0.000210</td>\n",
       "      <td>0.000185</td>\n",
       "      <td>0.000186</td>\n",
       "      <td>0.000134</td>\n",
       "      <td>0.000229</td>\n",
       "      <td>0.000102</td>\n",
       "      <td>0.000391</td>\n",
       "      <td>0.000161</td>\n",
       "      <td>0.000032</td>\n",
       "      <td>0.000077</td>\n",
       "      <td>0.000196</td>\n",
       "      <td>0.000052</td>\n",
       "      <td>0.000080</td>\n",
       "      <td>0.000221</td>\n",
       "      <td>0.000148</td>\n",
       "      <td>0.000171</td>\n",
       "      <td>0.000133</td>\n",
       "      <td>0.000091</td>\n",
       "      <td>0.000868</td>\n",
       "      <td>0.000036</td>\n",
       "      <td>0.000143</td>\n",
       "      <td>0.000201</td>\n",
       "      <td>0.922413</td>\n",
       "      <td>0.000119</td>\n",
       "      <td>0.000173</td>\n",
       "      <td>0.000127</td>\n",
       "      <td>0.000124</td>\n",
       "      <td>0.000162</td>\n",
       "      <td>0.000237</td>\n",
       "      <td>0.000137</td>\n",
       "      <td>0.000139</td>\n",
       "      <td>0.000161</td>\n",
       "      <td>0.000120</td>\n",
       "      <td>0.000049</td>\n",
       "      <td>0.000066</td>\n",
       "      <td>0.000198</td>\n",
       "      <td>0.000141</td>\n",
       "      <td>0.000105</td>\n",
       "      <td>0.000078</td>\n",
       "      <td>0.000124</td>\n",
       "      <td>0.000129</td>\n",
       "      <td>0.000117</td>\n",
       "      <td>0.000217</td>\n",
       "      <td>0.000235</td>\n",
       "      <td>0.000183</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000128</td>\n",
       "      <td>0.000245</td>\n",
       "      <td>0.000254</td>\n",
       "      <td>0.000144</td>\n",
       "      <td>0.000026</td>\n",
       "      <td>0.000168</td>\n",
       "      <td>0.000178</td>\n",
       "      <td>0.000263</td>\n",
       "      <td>0.000063</td>\n",
       "      <td>0.000192</td>\n",
       "      <td>0.000294</td>\n",
       "      <td>0.000148</td>\n",
       "      <td>0.000144</td>\n",
       "      <td>0.000124</td>\n",
       "      <td>0.000160</td>\n",
       "      <td>0.000160</td>\n",
       "      <td>0.000220</td>\n",
       "      <td>0.000217</td>\n",
       "      <td>0.000196</td>\n",
       "      <td>0.000306</td>\n",
       "      <td>0.000180</td>\n",
       "      <td>0.000101</td>\n",
       "      <td>0.000082</td>\n",
       "      <td>0.000132</td>\n",
       "      <td>0.000183</td>\n",
       "      <td>0.045154</td>\n",
       "      <td>0.000089</td>\n",
       "      <td>0.000130</td>\n",
       "      <td>0.000205</td>\n",
       "      <td>0.000096</td>\n",
       "      <td>0.000194</td>\n",
       "      <td>0.000085</td>\n",
       "      <td>0.000054</td>\n",
       "      <td>0.000075</td>\n",
       "      <td>0.000117</td>\n",
       "      <td>0.000142</td>\n",
       "      <td>0.000101</td>\n",
       "      <td>0.000084</td>\n",
       "      <td>0.000127</td>\n",
       "      <td>0.000119</td>\n",
       "      <td>0.000167</td>\n",
       "      <td>0.000159</td>\n",
       "      <td>0.000231</td>\n",
       "      <td>0.000155</td>\n",
       "      <td>0.000193</td>\n",
       "      <td>0.000149</td>\n",
       "      <td>0.000120</td>\n",
       "      <td>0.000330</td>\n",
       "      <td>0.000162</td>\n",
       "      <td>0.000162</td>\n",
       "      <td>0.000203</td>\n",
       "      <td>0.000242</td>\n",
       "      <td>0.000134</td>\n",
       "      <td>0.000212</td>\n",
       "      <td>0.000218</td>\n",
       "      <td>0.000351</td>\n",
       "      <td>0.000141</td>\n",
       "      <td>0.000152</td>\n",
       "      <td>0.000173</td>\n",
       "      <td>0.000112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.000046</td>\n",
       "      <td>0.000075</td>\n",
       "      <td>0.000154</td>\n",
       "      <td>0.000027</td>\n",
       "      <td>0.000063</td>\n",
       "      <td>0.000036</td>\n",
       "      <td>0.000068</td>\n",
       "      <td>0.000053</td>\n",
       "      <td>0.000053</td>\n",
       "      <td>0.000033</td>\n",
       "      <td>0.000059</td>\n",
       "      <td>0.000049</td>\n",
       "      <td>0.271895</td>\n",
       "      <td>0.000055</td>\n",
       "      <td>0.000051</td>\n",
       "      <td>0.000065</td>\n",
       "      <td>0.034911</td>\n",
       "      <td>0.000058</td>\n",
       "      <td>0.000042</td>\n",
       "      <td>0.000071</td>\n",
       "      <td>0.000032</td>\n",
       "      <td>0.007092</td>\n",
       "      <td>0.000050</td>\n",
       "      <td>0.000010</td>\n",
       "      <td>0.000024</td>\n",
       "      <td>0.000061</td>\n",
       "      <td>0.000016</td>\n",
       "      <td>0.000025</td>\n",
       "      <td>0.000068</td>\n",
       "      <td>0.000046</td>\n",
       "      <td>0.000053</td>\n",
       "      <td>0.000041</td>\n",
       "      <td>0.000028</td>\n",
       "      <td>0.000269</td>\n",
       "      <td>0.000011</td>\n",
       "      <td>0.000044</td>\n",
       "      <td>0.000062</td>\n",
       "      <td>0.000041</td>\n",
       "      <td>0.000037</td>\n",
       "      <td>0.000054</td>\n",
       "      <td>0.000039</td>\n",
       "      <td>0.000039</td>\n",
       "      <td>0.000050</td>\n",
       "      <td>0.000074</td>\n",
       "      <td>0.000043</td>\n",
       "      <td>0.000043</td>\n",
       "      <td>0.000050</td>\n",
       "      <td>0.000037</td>\n",
       "      <td>0.000015</td>\n",
       "      <td>0.000020</td>\n",
       "      <td>0.000061</td>\n",
       "      <td>0.000044</td>\n",
       "      <td>0.000032</td>\n",
       "      <td>0.000024</td>\n",
       "      <td>0.000038</td>\n",
       "      <td>0.000040</td>\n",
       "      <td>0.000036</td>\n",
       "      <td>0.000067</td>\n",
       "      <td>0.000073</td>\n",
       "      <td>0.000057</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000040</td>\n",
       "      <td>0.027959</td>\n",
       "      <td>0.000079</td>\n",
       "      <td>0.000045</td>\n",
       "      <td>0.000008</td>\n",
       "      <td>0.000052</td>\n",
       "      <td>0.000055</td>\n",
       "      <td>0.000082</td>\n",
       "      <td>0.000020</td>\n",
       "      <td>0.000059</td>\n",
       "      <td>0.000091</td>\n",
       "      <td>0.000046</td>\n",
       "      <td>0.000045</td>\n",
       "      <td>0.000038</td>\n",
       "      <td>0.000050</td>\n",
       "      <td>0.000050</td>\n",
       "      <td>0.000068</td>\n",
       "      <td>0.000067</td>\n",
       "      <td>0.000061</td>\n",
       "      <td>0.000095</td>\n",
       "      <td>0.000056</td>\n",
       "      <td>0.000031</td>\n",
       "      <td>0.000025</td>\n",
       "      <td>0.000041</td>\n",
       "      <td>0.000057</td>\n",
       "      <td>0.000051</td>\n",
       "      <td>0.000028</td>\n",
       "      <td>0.000040</td>\n",
       "      <td>0.000064</td>\n",
       "      <td>0.000030</td>\n",
       "      <td>0.000060</td>\n",
       "      <td>0.000026</td>\n",
       "      <td>0.000017</td>\n",
       "      <td>0.000023</td>\n",
       "      <td>0.000036</td>\n",
       "      <td>0.000044</td>\n",
       "      <td>0.000031</td>\n",
       "      <td>0.000026</td>\n",
       "      <td>0.000039</td>\n",
       "      <td>0.000037</td>\n",
       "      <td>0.000052</td>\n",
       "      <td>0.000049</td>\n",
       "      <td>0.000072</td>\n",
       "      <td>0.000048</td>\n",
       "      <td>0.000060</td>\n",
       "      <td>0.000046</td>\n",
       "      <td>0.097629</td>\n",
       "      <td>0.000102</td>\n",
       "      <td>0.000050</td>\n",
       "      <td>0.000050</td>\n",
       "      <td>0.000063</td>\n",
       "      <td>0.000075</td>\n",
       "      <td>0.000041</td>\n",
       "      <td>0.000066</td>\n",
       "      <td>0.000068</td>\n",
       "      <td>0.000109</td>\n",
       "      <td>0.000044</td>\n",
       "      <td>0.000047</td>\n",
       "      <td>0.000054</td>\n",
       "      <td>0.000035</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 200 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1         2         3         4         5         6  \\\n",
       "0  0.000196  0.000321  0.777824  0.000115  0.000269  0.000154  0.000291   \n",
       "1  0.000112  0.000184  0.000379  0.000066  0.000154  0.000088  0.000166   \n",
       "2  0.000158  0.000259  0.000534  0.000093  0.000217  0.000124  0.000234   \n",
       "3  0.000148  0.000242  0.000498  0.000086  0.000203  0.000116  0.000219   \n",
       "4  0.000046  0.000075  0.000154  0.000027  0.000063  0.000036  0.000068   \n",
       "\n",
       "          7         8         9        10        11        12        13  \\\n",
       "0  0.000227  0.000227  0.000143  0.000254  0.000210  0.000139  0.000235   \n",
       "1  0.000130  0.102773  0.000082  0.000146  0.000120  0.000080  0.000135   \n",
       "2  0.000184  0.000183  0.000116  0.000205  0.000169  0.000112  0.000190   \n",
       "3  0.000171  0.000171  0.000108  0.000191  0.000158  0.000105  0.000177   \n",
       "4  0.000053  0.000053  0.000033  0.000059  0.000049  0.271895  0.000055   \n",
       "\n",
       "         14        15        16        17        18        19        20  \\\n",
       "0  0.000221  0.000279  0.000245  0.000247  0.000178  0.000305  0.000135   \n",
       "1  0.000126  0.000160  0.000140  0.000141  0.000102  0.000174  0.000077   \n",
       "2  0.000178  0.000225  0.000198  0.000199  0.000144  0.000246  0.000109   \n",
       "3  0.000166  0.000210  0.000185  0.000186  0.000134  0.000229  0.000102   \n",
       "4  0.000051  0.000065  0.034911  0.000058  0.000042  0.000071  0.000032   \n",
       "\n",
       "         21        22        23        24        25        26        27  \\\n",
       "0  0.090193  0.000215  0.000042  0.000103  0.000260  0.000069  0.000106   \n",
       "1  0.000298  0.000123  0.000024  0.000059  0.000149  0.000040  0.000061   \n",
       "2  0.000420  0.000173  0.000034  0.000083  0.000210  0.000056  0.000086   \n",
       "3  0.000391  0.000161  0.000032  0.000077  0.000196  0.000052  0.000080   \n",
       "4  0.007092  0.000050  0.000010  0.000024  0.000061  0.000016  0.000025   \n",
       "\n",
       "         28        29        30        31        32        33        34  \\\n",
       "0  0.000294  0.000197  0.000227  0.000176  0.000121  0.031044  0.000048   \n",
       "1  0.000168  0.000113  0.000130  0.000101  0.000069  0.034874  0.000028   \n",
       "2  0.000237  0.000159  0.000183  0.000142  0.000098  0.025053  0.000039   \n",
       "3  0.000221  0.000148  0.000171  0.000133  0.000091  0.000868  0.000036   \n",
       "4  0.000068  0.000046  0.000053  0.000041  0.000028  0.000269  0.000011   \n",
       "\n",
       "         35        36        37        38        39        40        41  \\\n",
       "0  0.000190  0.000268  0.000177  0.000158  0.000230  0.000169  0.000165   \n",
       "1  0.000109  0.000153  0.000101  0.000091  0.000131  0.000097  0.000095   \n",
       "2  0.000154  0.000216  0.892676  0.000128  0.000185  0.000137  0.000133   \n",
       "3  0.000143  0.000201  0.922413  0.000119  0.000173  0.000127  0.000124   \n",
       "4  0.000044  0.000062  0.000041  0.000037  0.000054  0.000039  0.000039   \n",
       "\n",
       "         42        43        44        45        46        47        48  \\\n",
       "0  0.000215  0.000315  0.000183  0.000184  0.000214  0.000159  0.000065   \n",
       "1  0.000123  0.000180  0.000104  0.000106  0.000123  0.000091  0.000037   \n",
       "2  0.000173  0.000254  0.000147  0.000149  0.000173  0.000128  0.000052   \n",
       "3  0.000162  0.000237  0.000137  0.000139  0.000161  0.000120  0.000049   \n",
       "4  0.000050  0.000074  0.000043  0.000043  0.000050  0.000037  0.000015   \n",
       "\n",
       "         49        50        51        52        53        54        55  \\\n",
       "0  0.000088  0.000263  0.000187  0.000139  0.000104  0.000165  0.000172   \n",
       "1  0.000050  0.000151  0.000107  0.000080  0.000060  0.000094  0.000098   \n",
       "2  0.000071  0.000213  0.000151  0.000112  0.000084  0.000133  0.000139   \n",
       "3  0.000066  0.000198  0.000141  0.000105  0.000078  0.000124  0.000129   \n",
       "4  0.000020  0.000061  0.000044  0.000032  0.000024  0.000038  0.000040   \n",
       "\n",
       "         56        57        58        59    ...          140       141  \\\n",
       "0  0.000155  0.000288  0.000312  0.000243    ...     0.000170  0.000325   \n",
       "1  0.000089  0.000165  0.000178  0.000139    ...     0.000097  0.000186   \n",
       "2  0.000125  0.000233  0.000252  0.000196    ...     0.000137  0.000263   \n",
       "3  0.000117  0.000217  0.000235  0.000183    ...     0.000128  0.000245   \n",
       "4  0.000036  0.000067  0.000073  0.000057    ...     0.000040  0.027959   \n",
       "\n",
       "        142       143       144       145       146       147       148  \\\n",
       "0  0.000338  0.000191  0.000035  0.000223  0.000237  0.000350  0.000084   \n",
       "1  0.000193  0.000109  0.000020  0.000128  0.000135  0.000200  0.000048   \n",
       "2  0.000273  0.000154  0.000028  0.000180  0.000191  0.000282  0.000068   \n",
       "3  0.000254  0.000144  0.000026  0.000168  0.000178  0.000263  0.000063   \n",
       "4  0.000079  0.000045  0.000008  0.000052  0.000055  0.000082  0.000020   \n",
       "\n",
       "        149       150       151       152       153       154       155  \\\n",
       "0  0.000255  0.000391  0.000196  0.000191  0.000165  0.000213  0.000213   \n",
       "1  0.000146  0.000224  0.000112  0.000109  0.000094  0.000122  0.325158   \n",
       "2  0.000206  0.000316  0.000158  0.000154  0.000133  0.000172  0.000172   \n",
       "3  0.000192  0.000294  0.000148  0.000144  0.000124  0.000160  0.000160   \n",
       "4  0.000059  0.000091  0.000046  0.000045  0.000038  0.000050  0.000050   \n",
       "\n",
       "        156       157       158       159       160       161       162  \\\n",
       "0  0.000292  0.000288  0.000261  0.000406  0.000239  0.000134  0.000109   \n",
       "1  0.000167  0.000165  0.000149  0.000232  0.000137  0.000077  0.000062   \n",
       "2  0.000236  0.000232  0.000210  0.024450  0.000193  0.000108  0.000088   \n",
       "3  0.000220  0.000217  0.000196  0.000306  0.000180  0.000101  0.000082   \n",
       "4  0.000068  0.000067  0.000061  0.000095  0.000056  0.000031  0.000025   \n",
       "\n",
       "        163       164       165       166       167       168       169  \\\n",
       "0  0.000175  0.000244  0.000219  0.000119  0.000173  0.000272  0.000128   \n",
       "1  0.000100  0.000140  0.000125  0.000068  0.000099  0.000156  0.000073   \n",
       "2  0.000141  0.000197  0.000177  0.000096  0.000139  0.000220  0.000103   \n",
       "3  0.000132  0.000183  0.045154  0.000089  0.000130  0.000205  0.000096   \n",
       "4  0.000041  0.000057  0.000051  0.000028  0.000040  0.000064  0.000030   \n",
       "\n",
       "        170       171       172       173       174       175       176  \\\n",
       "0  0.000257  0.000112  0.000072  0.000100  0.000155  0.000189  0.000134   \n",
       "1  0.000147  0.000064  0.000041  0.000057  0.000089  0.000108  0.000077   \n",
       "2  0.000208  0.000091  0.000058  0.000081  0.000125  0.000153  0.000108   \n",
       "3  0.000194  0.000085  0.000054  0.000075  0.000117  0.000142  0.000101   \n",
       "4  0.000060  0.000026  0.000017  0.000023  0.000036  0.000044  0.000031   \n",
       "\n",
       "        177       178       179       180       181       182       183  \\\n",
       "0  0.000111  0.000168  0.000158  0.000222  0.000211  0.000307  0.000206   \n",
       "1  0.000064  0.000096  0.000090  0.000127  0.000121  0.000176  0.000118   \n",
       "2  0.000090  0.000136  0.000127  0.000179  0.000170  0.000248  0.000166   \n",
       "3  0.000084  0.000127  0.000119  0.000167  0.000159  0.000231  0.000155   \n",
       "4  0.000026  0.000039  0.000037  0.000052  0.000049  0.000072  0.000048   \n",
       "\n",
       "        184       185       186       187       188       189       190  \\\n",
       "0  0.000256  0.000198  0.000159  0.000438  0.000215  0.000215  0.000269   \n",
       "1  0.000147  0.000113  0.000091  0.000251  0.000123  0.000123  0.000154   \n",
       "2  0.000207  0.000159  0.000129  0.000354  0.000174  0.000174  0.000217   \n",
       "3  0.000193  0.000149  0.000120  0.000330  0.000162  0.000162  0.000203   \n",
       "4  0.000060  0.000046  0.097629  0.000102  0.000050  0.000050  0.000063   \n",
       "\n",
       "        191       192       193       194       195       196       197  \\\n",
       "0  0.000322  0.000177  0.000282  0.000290  0.000466  0.000188  0.000201   \n",
       "1  0.000184  0.051423  0.000161  0.000166  0.000267  0.000108  0.000115   \n",
       "2  0.000260  0.000143  0.000228  0.000234  0.000376  0.000152  0.000162   \n",
       "3  0.000242  0.000134  0.000212  0.000218  0.000351  0.000141  0.000152   \n",
       "4  0.000075  0.000041  0.000066  0.000068  0.000109  0.000044  0.000047   \n",
       "\n",
       "        198       199  \n",
       "0  0.000230  0.000148  \n",
       "1  0.000131  0.000085  \n",
       "2  0.000185  0.000120  \n",
       "3  0.000173  0.000112  \n",
       "4  0.000054  0.000035  \n",
       "\n",
       "[5 rows x 200 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print('Shape of dataframe:', df_doctopic.shape)\n",
    "df_doctopic.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2 Test Corpus"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**We could select an article or multiple atricles from the dataset we are using, or even any text from any where, as our test corpus. For instance, we will use below articale as test corpus:**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    ">**[Home New Zealand at War First World War Home front First World War homecomings](https://nzhistory.govt.nz/war/first-world-war-homecomings)**\n",
    ">\n",
    ">First World War homecomings\n",
    ">  \n",
    ">A soldier's story of battle\n",
    ">New Zealand troops began returning from overseas in small numbers in late 1914. The first arrivals, from Samoa, were followed by many more from Egypt and Turkey, and later from England and France. Most were sick or wounded, but some were sent back because of misconduct. More troops returned home from early 1918, in line with the NZEF's scheme for demobilisation. Most were 'low category men' such as those deemed permanently unfit, who were sent home to free up hospital beds for the heavy casualties that were anticipated. Some fit men who had served for several years were sent home on furlough (leave of absence).\n",
    ">\n",
    ">[Image: When the Boys Come Home music]  \n",
    ">'When the Boys Come Home'\n",
    ">\n",
    ">By the time of the armistice with Germany in November 1918, more than 24,000 men had returned home. But more than 56,000 remained overseas or were at sea. The Defence Department arranged the return of the vast majority of these troops during 1919. It also transported thousands more people: the wives and children of soldiers, civilian war workers, and New Zealanders who had served in the British army or navy.\n",
    ">\n",
    ">..."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Save the article in** `.tsv` **file to infer topic model by MALLET.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>title</th>\n",
       "      <th>content</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>First game of rugby played in NZ</td>\n",
       "      <td>Home New Zealand at War First World War Home front First World War homecomings  First World War homecomings  A soldi...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   id                             title  \\\n",
       "0   0  First game of rugby played in NZ   \n",
       "\n",
       "                                                                                                                   content  \n",
       "0  Home New Zealand at War First World War Home front First World War homecomings  First World War homecomings  A soldi...  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "with open('../temp/test_corpus.txt') as f:  \n",
    "    text = ' '.join(line.strip() for line in f)\n",
    "\n",
    "df = pd.DataFrame(data={'id':[0], 'title':['First game of rugby played in NZ'], 'content':[text]})\n",
    "df = df[['id', 'title', 'content']]\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "path = r'../data/dataset/similarity/corpus.tsv'\n",
    "df.to_csv(path, sep='\\t', header=False, index=False, encoding='utf-8')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Then we will use this small corpus to find similar articles in the papers past dataset.**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3 Infer Topic Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**We use MALLET to infer the corpus to get doc-topic matrix:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "InputFile=../data/dataset/similarity/corpus.tsv\n",
      "OutputDir=../models/similarity\n",
      "Process=infer\n",
      "TrainDir=../models/train\n",
      "Inferencer=../models/train/inferencer.model\n",
      "CORES=12\n",
      "SEED1=10839\n",
      "SEED2=23012\n",
      "TOPICS=200\n",
      "ITERATION=2000\n",
      "INTERVAL=40\n",
      "BURNIN=300\n",
      "IDFMIN=0\n",
      "IDFMAX=8\n",
      "20:32:44 :: Start import dataset...\n",
      "Import file already exist, nothing to do.\n",
      "20:32:44 :: Imported.\n",
      "Pruned model already exist, nothing to do.\n",
      "20:32:44 :: Pruned.\n",
      "20:32:44 :: Start infering dataset...\n",
      "20:33:00 :: Inferred.\n"
     ]
    }
   ],
   "source": [
    "%%bash\n",
    "\n",
    "bash ../5-modeling/model.sh -i '../data/dataset/similarity/corpus.tsv' -o  '../models/similarity' -p 'infer'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape of dataset: (1, 200)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>10</th>\n",
       "      <th>11</th>\n",
       "      <th>12</th>\n",
       "      <th>13</th>\n",
       "      <th>14</th>\n",
       "      <th>15</th>\n",
       "      <th>16</th>\n",
       "      <th>17</th>\n",
       "      <th>18</th>\n",
       "      <th>19</th>\n",
       "      <th>20</th>\n",
       "      <th>21</th>\n",
       "      <th>22</th>\n",
       "      <th>23</th>\n",
       "      <th>24</th>\n",
       "      <th>25</th>\n",
       "      <th>26</th>\n",
       "      <th>27</th>\n",
       "      <th>28</th>\n",
       "      <th>29</th>\n",
       "      <th>30</th>\n",
       "      <th>31</th>\n",
       "      <th>32</th>\n",
       "      <th>33</th>\n",
       "      <th>34</th>\n",
       "      <th>35</th>\n",
       "      <th>36</th>\n",
       "      <th>37</th>\n",
       "      <th>38</th>\n",
       "      <th>39</th>\n",
       "      <th>40</th>\n",
       "      <th>41</th>\n",
       "      <th>42</th>\n",
       "      <th>43</th>\n",
       "      <th>44</th>\n",
       "      <th>45</th>\n",
       "      <th>46</th>\n",
       "      <th>47</th>\n",
       "      <th>48</th>\n",
       "      <th>49</th>\n",
       "      <th>50</th>\n",
       "      <th>51</th>\n",
       "      <th>52</th>\n",
       "      <th>53</th>\n",
       "      <th>54</th>\n",
       "      <th>55</th>\n",
       "      <th>56</th>\n",
       "      <th>57</th>\n",
       "      <th>58</th>\n",
       "      <th>59</th>\n",
       "      <th>...</th>\n",
       "      <th>140</th>\n",
       "      <th>141</th>\n",
       "      <th>142</th>\n",
       "      <th>143</th>\n",
       "      <th>144</th>\n",
       "      <th>145</th>\n",
       "      <th>146</th>\n",
       "      <th>147</th>\n",
       "      <th>148</th>\n",
       "      <th>149</th>\n",
       "      <th>150</th>\n",
       "      <th>151</th>\n",
       "      <th>152</th>\n",
       "      <th>153</th>\n",
       "      <th>154</th>\n",
       "      <th>155</th>\n",
       "      <th>156</th>\n",
       "      <th>157</th>\n",
       "      <th>158</th>\n",
       "      <th>159</th>\n",
       "      <th>160</th>\n",
       "      <th>161</th>\n",
       "      <th>162</th>\n",
       "      <th>163</th>\n",
       "      <th>164</th>\n",
       "      <th>165</th>\n",
       "      <th>166</th>\n",
       "      <th>167</th>\n",
       "      <th>168</th>\n",
       "      <th>169</th>\n",
       "      <th>170</th>\n",
       "      <th>171</th>\n",
       "      <th>172</th>\n",
       "      <th>173</th>\n",
       "      <th>174</th>\n",
       "      <th>175</th>\n",
       "      <th>176</th>\n",
       "      <th>177</th>\n",
       "      <th>178</th>\n",
       "      <th>179</th>\n",
       "      <th>180</th>\n",
       "      <th>181</th>\n",
       "      <th>182</th>\n",
       "      <th>183</th>\n",
       "      <th>184</th>\n",
       "      <th>185</th>\n",
       "      <th>186</th>\n",
       "      <th>187</th>\n",
       "      <th>188</th>\n",
       "      <th>189</th>\n",
       "      <th>190</th>\n",
       "      <th>191</th>\n",
       "      <th>192</th>\n",
       "      <th>193</th>\n",
       "      <th>194</th>\n",
       "      <th>195</th>\n",
       "      <th>196</th>\n",
       "      <th>197</th>\n",
       "      <th>198</th>\n",
       "      <th>199</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.000003</td>\n",
       "      <td>0.000005</td>\n",
       "      <td>0.000868</td>\n",
       "      <td>0.000002</td>\n",
       "      <td>0.000004</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>0.000016</td>\n",
       "      <td>0.000108</td>\n",
       "      <td>0.05336</td>\n",
       "      <td>0.000014</td>\n",
       "      <td>0.052967</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>0.000002</td>\n",
       "      <td>0.000108</td>\n",
       "      <td>0.00005</td>\n",
       "      <td>0.000074</td>\n",
       "      <td>0.000213</td>\n",
       "      <td>0.000004</td>\n",
       "      <td>0.000223</td>\n",
       "      <td>0.025687</td>\n",
       "      <td>0.000049</td>\n",
       "      <td>0.148629</td>\n",
       "      <td>0.000038</td>\n",
       "      <td>6.828897e-07</td>\n",
       "      <td>0.00021</td>\n",
       "      <td>0.003261</td>\n",
       "      <td>0.000187</td>\n",
       "      <td>0.000083</td>\n",
       "      <td>0.000492</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>0.000641</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>0.000002</td>\n",
       "      <td>0.093799</td>\n",
       "      <td>7.866837e-07</td>\n",
       "      <td>0.009193</td>\n",
       "      <td>0.006877</td>\n",
       "      <td>0.015706</td>\n",
       "      <td>0.000165</td>\n",
       "      <td>0.000085</td>\n",
       "      <td>0.001023</td>\n",
       "      <td>0.000582</td>\n",
       "      <td>0.000328</td>\n",
       "      <td>0.000028</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>0.000015</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>0.000802</td>\n",
       "      <td>0.000001</td>\n",
       "      <td>0.000001</td>\n",
       "      <td>0.000027</td>\n",
       "      <td>0.000096</td>\n",
       "      <td>0.019959</td>\n",
       "      <td>0.000002</td>\n",
       "      <td>0.0002</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>0.000005</td>\n",
       "      <td>0.000005</td>\n",
       "      <td>0.000004</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000084</td>\n",
       "      <td>0.000052</td>\n",
       "      <td>0.001327</td>\n",
       "      <td>0.000061</td>\n",
       "      <td>0.020062</td>\n",
       "      <td>0.000108</td>\n",
       "      <td>0.001847</td>\n",
       "      <td>0.000017</td>\n",
       "      <td>0.000222</td>\n",
       "      <td>0.000004</td>\n",
       "      <td>0.029767</td>\n",
       "      <td>0.003213</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>0.00793</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>0.001894</td>\n",
       "      <td>0.000097</td>\n",
       "      <td>0.000143</td>\n",
       "      <td>0.000007</td>\n",
       "      <td>0.000027</td>\n",
       "      <td>0.001022</td>\n",
       "      <td>0.000002</td>\n",
       "      <td>0.078137</td>\n",
       "      <td>0.000004</td>\n",
       "      <td>0.027957</td>\n",
       "      <td>0.000002</td>\n",
       "      <td>0.000165</td>\n",
       "      <td>0.000028</td>\n",
       "      <td>0.000002</td>\n",
       "      <td>0.022035</td>\n",
       "      <td>0.00604</td>\n",
       "      <td>0.000245</td>\n",
       "      <td>0.000002</td>\n",
       "      <td>0.000118</td>\n",
       "      <td>0.007652</td>\n",
       "      <td>0.017224</td>\n",
       "      <td>0.000048</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>0.000038</td>\n",
       "      <td>0.022533</td>\n",
       "      <td>0.000017</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>0.00289</td>\n",
       "      <td>0.000003</td>\n",
       "      <td>0.000524</td>\n",
       "      <td>0.000065</td>\n",
       "      <td>0.000004</td>\n",
       "      <td>0.000004</td>\n",
       "      <td>0.000016</td>\n",
       "      <td>0.000017</td>\n",
       "      <td>0.000026</td>\n",
       "      <td>0.012915</td>\n",
       "      <td>0.000039</td>\n",
       "      <td>0.000008</td>\n",
       "      <td>0.000084</td>\n",
       "      <td>0.000061</td>\n",
       "      <td>0.000015</td>\n",
       "      <td>0.000246</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1 rows × 200 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1         2         3         4         5         6  \\\n",
       "0  0.000003  0.000005  0.000868  0.000002  0.000004  0.000003  0.000016   \n",
       "\n",
       "          7        8         9        10        11        12        13  \\\n",
       "0  0.000108  0.05336  0.000014  0.052967  0.000003  0.000002  0.000108   \n",
       "\n",
       "        14        15        16        17        18        19        20  \\\n",
       "0  0.00005  0.000074  0.000213  0.000004  0.000223  0.025687  0.000049   \n",
       "\n",
       "         21        22            23       24        25        26        27  \\\n",
       "0  0.148629  0.000038  6.828897e-07  0.00021  0.003261  0.000187  0.000083   \n",
       "\n",
       "         28        29        30        31        32        33            34  \\\n",
       "0  0.000492  0.000003  0.000641  0.000003  0.000002  0.093799  7.866837e-07   \n",
       "\n",
       "         35        36        37        38        39        40        41  \\\n",
       "0  0.009193  0.006877  0.015706  0.000165  0.000085  0.001023  0.000582   \n",
       "\n",
       "         42        43        44        45        46        47        48  \\\n",
       "0  0.000328  0.000028  0.000003  0.000015  0.000003  0.000802  0.000001   \n",
       "\n",
       "         49        50        51        52        53      54        55  \\\n",
       "0  0.000001  0.000027  0.000096  0.019959  0.000002  0.0002  0.000003   \n",
       "\n",
       "         56        57        58        59    ...          140       141  \\\n",
       "0  0.000003  0.000005  0.000005  0.000004    ...     0.000084  0.000052   \n",
       "\n",
       "        142       143       144       145       146       147       148  \\\n",
       "0  0.001327  0.000061  0.020062  0.000108  0.001847  0.000017  0.000222   \n",
       "\n",
       "        149       150       151       152       153      154       155  \\\n",
       "0  0.000004  0.029767  0.003213  0.000003  0.000003  0.00793  0.000003   \n",
       "\n",
       "        156       157       158       159       160       161       162  \\\n",
       "0  0.001894  0.000097  0.000143  0.000007  0.000027  0.001022  0.000002   \n",
       "\n",
       "        163       164       165       166       167       168       169  \\\n",
       "0  0.078137  0.000004  0.027957  0.000002  0.000165  0.000028  0.000002   \n",
       "\n",
       "        170      171       172       173       174       175       176  \\\n",
       "0  0.022035  0.00604  0.000245  0.000002  0.000118  0.007652  0.017224   \n",
       "\n",
       "        177       178       179       180       181       182       183  \\\n",
       "0  0.000048  0.000003  0.000003  0.000038  0.022533  0.000017  0.000003   \n",
       "\n",
       "       184       185       186       187       188       189       190  \\\n",
       "0  0.00289  0.000003  0.000524  0.000065  0.000004  0.000004  0.000016   \n",
       "\n",
       "        191       192       193       194       195       196       197  \\\n",
       "0  0.000017  0.000026  0.012915  0.000039  0.000008  0.000084  0.000061   \n",
       "\n",
       "        198       199  \n",
       "0  0.000015  0.000246  \n",
       "\n",
       "[1 rows x 200 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "path = r'../models/similarity/docTopicsInfer.txt'\n",
    "\n",
    "df_doctopic_test = (load_doctopic(path, topics_shape[0], spark)).drop('index').drop('id')\n",
    "df_doctopic_test = df_doctopic_test.toPandas()\n",
    "\n",
    "print('Shape of dataset:', df_doctopic_test.shape)\n",
    "df_doctopic_test.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**List the highest weight 10 topics:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>weight</th>\n",
       "      <th>topic</th>\n",
       "      <th>keywords</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0.148629</td>\n",
       "      <td>21</td>\n",
       "      <td>public time present question people great fact matter good made country case men position government doubt opinion m...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>0.093799</td>\n",
       "      <td>33</td>\n",
       "      <td>tin ill oil lit mil day zealand ton tie hat aro fur aid tins ail end man hut til time</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>163</th>\n",
       "      <td>0.078137</td>\n",
       "      <td>163</td>\n",
       "      <td>late son daughter funeral residence aged years mrs wife street cemetery john leave beloved friends wellington rev wi...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.053360</td>\n",
       "      <td>8</td>\n",
       "      <td>goods books fancy christmas stock cases large sets presents xmas assortment cards prices toys variety boxes statione...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.052967</td>\n",
       "      <td>10</td>\n",
       "      <td>british empire britain imperial conference great new_zealand war sir canada australia colonies government london dom...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       weight  topic  \\\n",
       "21   0.148629     21   \n",
       "33   0.093799     33   \n",
       "163  0.078137    163   \n",
       "8    0.053360      8   \n",
       "10   0.052967     10   \n",
       "\n",
       "                                                                                                                    keywords  \n",
       "21   public time present question people great fact matter good made country case men position government doubt opinion m...  \n",
       "33                                    tin ill oil lit mil day zealand ton tie hat aro fur aid tins ail end man hut til time   \n",
       "163  late son daughter funeral residence aged years mrs wife street cemetery john leave beloved friends wellington rev wi...  \n",
       "8    goods books fancy christmas stock cases large sets presents xmas assortment cards prices toys variety boxes statione...  \n",
       "10   british empire britain imperial conference great new_zealand war sir canada australia colonies government london dom...  "
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_plt = df_doctopic_test.T\n",
    "df_plt.columns = ['weight']\n",
    "df_plt = df_plt.reset_index(drop=True)\n",
    "df_plt[['topic', 'keywords']] = df_topics[['topic_', 'keywords_']]\n",
    "df_plt.sort_values(by='weight', ascending=False).head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Plot Weight:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1040x480 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot\n",
    "fig, axes = plt.subplots(figsize=(13, 6), dpi=dpi)\n",
    "df_plt.plot(ax=axes, \n",
    "            kind='bar', \n",
    "            x='topic', \n",
    "            y='weight', \n",
    "            legend=False)\n",
    "axes.set_title('Topic Distribution of the Test Corpus', fontdict=dict(size=16))\n",
    "axes.set_ylabel('Topic Weight')\n",
    "axes.set_xlabel('Topic')\n",
    "for label in axes.get_xticklabels()[::2]:\n",
    "    label.set_visible(False)\n",
    "fig.tight_layout()\n",
    "plt.xticks(rotation='90')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4 Jensen-Shannon Divergence "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    ">**Similarity query**\n",
    ">\n",
    ">Now that we have a topic distribution for a new unseen document, let's say we wanted to find the most similar documents in the corpus. We can do this by comparing the topic distribution of the new document to all the topic distributions of the documents in the corpus. We use the [Jensen-Shannon distance](https://en.wikipedia.org/wiki/Jensen%E2%80%93Shannon_divergence) metric to find the most similar documents.\n",
    ">\n",
    ">What the Jensen-Shannon distance tells us, is which documents are statisically \"closer\" (and therefore more similar), by comparing the divergence of their distributions. Jensen-Shannon is symmetric, unlike Kullback-Leibler on which the formula is based. This is good, because we want the similarity between documents A and B to be the same as the similarity between B and A.\n",
    ">\n",
    ">The formula is described below.\n",
    ">\n",
    ">For discrete distirbutions $P$ and $Q$, the Jensen-Shannon divergence, $JSD$ is defined as\n",
    ">\n",
    ">$$JSD\\left(P||Q\\right) = \\frac{1}{2}D\\left(P||M\\right)+\\frac{1}{2}D\\left(Q||M\\right)$$\n",
    ">\n",
    ">where $M = \\frac{1}{2}\\left(P+Q\\right)$\n",
    ">\n",
    ">and $D$ is the Kullback-Leibler divergence\n",
    ">\n",
    ">$$D\\left(P||Q\\right) = \\sum_iP(i)\\log\\left(\\frac{P(i)}{Q(i)}\\right)$$\n",
    ">\n",
    ">$$\\Rightarrow JSD\\left(P||Q\\right) = \\frac{1}{2}\\sum_i\n",
    ">\\left[\n",
    ">P(i)\\log\\left(\\frac{P(i)}{\\frac{1}{2}\\left(P(i)+Q(i)\\right)}\\right)\n",
    ">+\n",
    ">Q(i)\\log\\left(\\frac{Q(i)}{\\frac{1}{2}\\left(P(i)+Q(i)\\right)}\\right)\n",
    ">\\right]$$\n",
    ">\n",
    ">The square root of the Jensen-Shannon divergence is the Jensen-Shannon Distance: $\\sqrt{JSD\\left ( P||Q\\right )}$\n",
    ">\n",
    ">**The smaller the Jensen-Shannon Distance, the more similar two distributions are (and in our case, the more similar any 2 documents are)**\n",
    ">\n",
    ">We can use the scipy implementation of entropy to do this. Entropy calculates the KL divergence.\n",
    ">\n",
    ">We got a dense matrix - doc-topic matrix - from MALLET, of size $M\\times K$ where $M$ is the number of documents and $K$ is the number of topics.\n",
    ">\n",
    ">We will comparing the new unseen document above to the entire train corpora to find which one is most similar to the new document.\n",
    "\n",
    "**How are we doing that? We have the new documents LDA topic distribution in stored as varibale new doc-topic matrix, and we have the entire corpus of documents topic distributions stored in the dense matrix doc-topic matrix. Then we pass each row of doc-topic matrix through the Jensen-Shannon function above as the Q distribution, while the P distribution remains static as new doc-topic matrix. Then we get the smallest distances and their corresponding index in the array, which we can pass to the train corpora to print out the most similar documents.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "def jensen_shannon(query, matrix):\n",
    "    \"\"\"\n",
    "    This function implements a Jensen-Shannon similarity\n",
    "    between the input query (an LDA topic distribution for a document)\n",
    "    and the entire corpus of topic distributions.\n",
    "    It returns an array of length M where M is the number of documents in the corpus\n",
    "    \"\"\"\n",
    "    # lets keep with the p,q notation above\n",
    "    p = query.T.values\n",
    "    q = matrix.T.values\n",
    "    m = 0.5*(p + q)\n",
    "    return (0.5*(entropy(p,m) + entropy(q,m)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_similar(query,matrix,k=5):\n",
    "    \"\"\"\n",
    "    This function implements the Jensen-Shannon distance above\n",
    "    and retruns the top k indices of the smallest jensen shannon distances\n",
    "    \"\"\"\n",
    "    sims = jensen_shannon(query,matrix) # list of jensen shannon distances\n",
    "    #print(sims)\n",
    "    return sims.argsort()[:k] # the top k positional index of the smallest Jensen Shannon distances"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The most similar 5 documents row index: [  69049 2724005 1142500    8684 2839839]\n"
     ]
    }
   ],
   "source": [
    "list_sim = get_similar(df_doctopic_test,df_doctopic)\n",
    "\n",
    "print('The most similar 5 documents row index:',list_sim)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5 Recommendation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Based on the Jensen-Shannon Divergence, we calculated similarity list, and use the list to recommend similar documents.**\n",
    "\n",
    "**Print meta information:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>publisher</th>\n",
       "      <th>region</th>\n",
       "      <th>date</th>\n",
       "      <th>ads</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8684</th>\n",
       "      <td>1929443</td>\n",
       "      <td>Bay Of Plenty Times</td>\n",
       "      <td>Bay of Plenty</td>\n",
       "      <td>1879-05-10</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>69049</th>\n",
       "      <td>2421886</td>\n",
       "      <td>Daily Southern Cross</td>\n",
       "      <td>Auckland</td>\n",
       "      <td>1861-05-17</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1142500</th>\n",
       "      <td>12564083</td>\n",
       "      <td>Star</td>\n",
       "      <td>Canterbury</td>\n",
       "      <td>1893-07-29</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2724005</th>\n",
       "      <td>28113158</td>\n",
       "      <td>Auckland Star</td>\n",
       "      <td>Auckland</td>\n",
       "      <td>1887-12-23</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2839839</th>\n",
       "      <td>29000415</td>\n",
       "      <td>Daily Telegraph</td>\n",
       "      <td>Hawke's Bay</td>\n",
       "      <td>1888-11-17</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               id             publisher         region        date    ads\n",
       "8684      1929443   Bay Of Plenty Times  Bay of Plenty  1879-05-10  False\n",
       "69049     2421886  Daily Southern Cross       Auckland  1861-05-17  False\n",
       "1142500  12564083                  Star     Canterbury  1893-07-29  False\n",
       "2724005  28113158         Auckland Star       Auckland  1887-12-23  False\n",
       "2839839  29000415       Daily Telegraph    Hawke's Bay  1888-11-17  False"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#df_similar = df_clean[df_clean.index.isin(most_sim_list)]\n",
    "df_similar = df_meta[df_meta.index.isin(list_sim)]\n",
    "df_similar.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Recommendation 1:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# linux line number is based on 0\n",
    "line = list_sim[0] +1 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "print the 69050 document:\n",
      "2421886\tTHE TIMES ON ITALIAN UNITY.\t\"THE TIMES ON ITALIAN UNITY.There are moments in the history of nations of an importance so transcendant and so overpowering that they combine and lesumo Ih themselves the wrongs, the agonies, and the Lopes of centuries, and raise men, ia the consciousness of a common sentiment and the transport of an ii repressible enthusiasm, far above the cares and interests, the small joys and paltry sorrows, of ordinary lite. Such moments are the poetry of histoiy, and, though tbe ideal which hah for the moment displaced the actual is fleeting and transitoiy, andthough the sordid caies of oidiuary existence once more a«seit their empire of a mind for a time laised above them, they stand like beacons in the dieaiy annals of mankind, and attiact the genius of all ages to study and understand them, Just as America has achieved the inglorious exploit of tearing heisclf to pieces aud pounding into dust the solid gronite of her institutions, Italy, after a tiialofa length and seventy totally unexampled in the annals of any other people, has at length achieved in a gieat measme the glorious exploit of Italian uuity. Tt lequired six huudred years of incessant bloodshed and conquest before Rome succeeded in thoroughly harmonizing and uniting the Italian Peninsula under her government; and scaicely had this mighty feat been achieved before the unity of the co'Kjueiing nation was lost iv the general servitude uh'ch equalued and overspread the whoie civilized woi-ld For a shot t tune after the fall of the Wntein Empiie, Ita'y was united a« it had been just befoie its use, but from that tune to the pre.-ent — that is, for ncaily fouiteen centuiies — Italy has sought and sighed for union in vain. She has successively fallen under the dominion of almost eveiy tribe that dwelt between the Wall of China and the Western coast of France : Groths, Vandals, ll ana, Lombards, Greeks, Fianks Siraceus, IS 01 mans, French, Spaniards,andGermms have all in turn wasted her fair inheritance, but none have attained to the glory of reuniting liei under a single sceptre That honour has been reserved for a Pimce who n.ukus no pictence to brilliant or extraordinary talent, but •who has shown himself a true patriot and a biave soldier True, the inheritance is not eovnp'ete Rome, heiself the ancient seat of empiie; Verona, the noble capital of Theodoric ; and Venice, which defied Attila, aud conquered Constantinople, still remain to be included ; bat these deficiencies, great as they are, serve only to mark how much has been gained, and to louse a not unieasonable hope that their acquisition will not be long defeired Well may the ancient capital of the House of Savoy bo proud of the distinction of being the fiist spot of Italian ground on which an Itilian Pailiament has met Well may she illuminate her streets and rend the air with her acclamations, fur she has not only the joy of beholding the resunection of Italy, but the consciousness that she and her sons ha\\\\e been the main actors in this gieat and good work. Whatever may be her fate to morrow, though she can haidly expect to maintain in a country where every part is alike free and happy the same rapid advance in population and wealth as when she was in Italy the land of Goshen, where alonp theie was light Tuna has every right to <he triumph of the present moment, aud to the respectful homage of those who from every quarter of renovated Italy have come to lay the foundations of a State which is hence fjrth to take its place among the Powers of Europe. The building in which the Italian Parliament meets seems to be scarcely worthy of the solemn occasion ; but a King who trusts, aid not in vain, th.it he la to be enthioned in the hearts of his people, may easily disregard any imperfections of an angement oc decoration, and au assembly winch con^iits of the lank and virtue, tbe wisdom aud modeiation of the whole country, fieely chosen by an entire nation and delegate i to lay broad and deep the foundations of her hbeity and happiness, could acquire no additional lustre were it assembled in the most gorgeous apartment of the Tuileries or the Kscurial How different from the calm and A lendly recognitions exchanged between our ,j vii representatives in the first days after the rjcess must be the feelings of the oralois, statesmen, soldiers, and martyrs of this new Italian Parliament, who beho d for the first time fiose in who' -*o deeds or sufferings they have felt for many jears so deep and absorbing an interest ! ilow electric must be the influence of those members on each other, consisting, as they do, almost entirely of men tried in every species of adversity, and distinguished in every department of intellectual activity, when to achieve distinction was to incur suspicion ! The speech of the King of Italy was in every respect worthy of the occasion. It soared above the homely common-place to which an English Pailiament is accustomed, without the affected obscuiity, the inflation, or the self-laudation which disfigures the orations of some other Royal personages The King aimed straight at the nuik, and hit it. Italy, he tells his Parliament, confide'? in thsir wisdom, and tiuits to t'lem her organization Tint oiganization s to consist of unity of Government, with a decentralized Administration. It is impossible t ) express more precisely the problem that has to be solved. Administrative centralization like that of France would infallibly provoke re-action and disruption by its outrage on the sentiment oi local pride and patriotism, while a Government which delegated the supreme power to lo al bodies must speedily perish through its own weakness. \\\"With qi eat good tnslc and simplicity the King notices the fact that the Envoy of France is not present on that sol emu occasion, but reminds his audience of the ola'nns on their gratitude established by Magenta and Solfeiino, and assures Fiance that Italy feels for her a friendship which cannot be dissolved We do not envy the Emporor of | Fienoh the fephngs with wh oh lie must icad , this, just and gentle leproof, nor tho reflection which must quobs his mind, of how much he has done to forfeit tho gratitude and affection of a nation which he had justly earned for himself byservices of such unexampled magnitude We in England* have good reason to be proud that our name \\\\(ms been thought worthy to be mentioned on theraay which inaugurates Italian liberty. We iMvcvbsd no blood, we have spent no treasure, but wcr^trve not wavered in our devotion to the principle of non-intervention, and we receive with sincere good will the assurance that the part which we have played dm ing the last two years will be held in imperishable le.nembrance. Even more than with the notice of England are we gratified with the wise and moderate tone of the King's address lie leminds his auditois that he risked his Ciowu for Italy, but that there is a tune to wait as well as a time to daro The destinies of a nation are not to be trifled with, aud Italy is strong enough to be pmdent These words satisfy us that it will not be by any nsh enterprise on tho part of the Sardinian Government that Austria will bo furnished with an excuse for an attempt to conceal her bankrupt exchequer aud her disafteeted empiie behind the smoke of gunpowder. The quoriel, if there is to be one, must be of her own seeking, for Italy is well aware that time fights for her, aud that eveiy month, while it consolidates her own power, strengthens her army, and incicases her confidence in herself, leaves the Empire of Austria moie hopelessly insolvent than it found her, and brings steadily nearer and nearer the rnomentof afrightfulbut inevitablecatastiophc. Italian unity has been inaugurated in a manner in every way worthy of the first solemn homy of a great nation, and we turn fiom the contemplation ot the scene with the sinceie wish, Esto Perpetua.Aeryted Bueyd —Within the last yeai 01 two it has occ.ined to a physician, Dr. Daughsh, that by mechamcal contrivance the piue an can be pissed into the dough, and tint floui unaltued by feimentation, untouched by any chemicil, unpolluted even by the touch of any hand, can be made into a peifect foim of spongy bread Having developed hia plan fully, he took out a patent, and aheady at Poitsmouth and at Dockhead, in Beimondsey, extensive faetones are engaged hi the production of an \\\"aeiated biead,\\\" which, as to its substance, is, we believe, bread made perfect, though it is possible that there may be hereaftei devel oped a less costly way of making it The patent is worked \\\\\\\\ holly by bteam machinery, of winch «c can not attempt to e\\\\plam all the ingenious refinements The main principle is easily to be uuderotood. Aceoid ing to the way usually adopted m pioducing the same gas foi soda watei, carbonic aeul is foimed m a laige receiver, fai away fiom the dough Thence it is foiced into a gieat copper eyhndei, containing watei, fixed over the mixing vessel At a high piessure, which is maintained also by the foicmg of the same gas within the mixing \\\\essel, the w xtei in the eyhndei is supusalmatcd \\\\uth gas—is made, m fact, into soda watei free fiom soda In that state it is then allowed to flow throug'i a pipe ovei the due lehtivepiopoitions of flour and salt, imdei the highly condensed atmospheie of the closed mixei The mixei is a hollow globe of cast 11011, in which thenon aims aie made to ievolve on an axis turned bv the steam engine The gab l emains fixed, still under presume, m the -watei In three 01 four minutes, or moie, according to the quality of the floui, the mixtme of the flour with the soda-watei is com plete The paste then passes out through a tube gi \\\\ dually widening, and the gas expands m eveiy pore of the dough, not breaking out of the decomposed Hour, but out of the water, as the piessuie is removed The dough instantly rises as it passes into the tins, or wooden measuics, which a boy holds under the spout, cutting off the mpftsmo of each loaf as it descends, and immediately placing it on the edge of the oven, which is on the other side of linn The flooi of the oven is nu endless chain, revolving on two drums, of which the pace is legulated in accordance with the size and chalacter of the bieal baked The loaves placed on one edge of the oven immediately begin to tiavel through its regulated heat, and in due time aie turned out exactly baked upon the other side, close to the open door, at which caits wait to carry the loaves to the shopkeeper. Until tho biead is baked not a hand touches it An hour and a half is lime enough foi the conveision, by this process, of a sack of flour into baked loaves, peifectly spongy, and with the nutritive elements of the floui wholly untouched. In the oidinary piocess four 01 five houis aie lequired foi the mere raising of the sponge Tins piolonged action of the warmth and moisture upon many kinds of flour —as floui fiom wheat gatheied m wet seasons—otherwise wholesome, changes the staichy mattci mto dextrine, and aftm all pioducos bread daik colouied and sodden. It is to collect so gieat an occasion of un certainty and loss, which has always prevented capi tahsts fiom embarking in the b iking trade, that alum has been used The rapidity of the new aeiating pio ctss wholly avoids this nsk; the result nevei is uncertain, and good bread can be made of flour otheiwisC almobt useless, to the bakei. Tim unfermented, 01, as it is pioperly called, the aeiated biead, made accouling to Di. Dauglioh'h patent, being entnely fiee fiom the acid which is always necessarily piesont in feimentcd biead, has been found actually curative in that numeious class of diseases which result fiom acid secietions 01 an acid state of the blood r l Ins freedom fiom acid causes the bread at fiist to appear somewhat insipid, but it soon asserts it 3 value. One of the most eminent of our physicians kept a loaf of it for a foitnight, and then caused it to appsai- at hi 3 bieakfast table with a baker's loaf of the preceding day 'lhc unfeimented lo if, old as it was, appeared to be the fresher of the two Experience has shown that working men who u-.e the aeiated bread eat more of it —sometime^ oven half as much again —making heaity breakfasts, and being at dinner-time loss hungiy foi meat At Uuj's Hospital the patients have always their quantity of biead by weight duly prescubed. Fermented b.ead, of the best quality, is used, m ide in the hospital, to insuie its being good. Of the quantity supplied to the waids there is gatheied eveiy day a large surplus winch the patients have been unable to consume For about two months two of the waids were supplied, by way of eKpenmont, with aerated biead m the usual quantities One remaikablo result was that fiom these wauls theie was never aiiy buiplus to be collected The sick stomachs never turned against it The use of aeiated bread tends lather, theiefoie, to the inciease of the baker's, and the deciease of the butchei's bill. Its actual price is that of ordinaly biead, but, as its manufacture demands costly machinery, it can be sold only in the shops of bakeis 01 coin dealers as books aie sold by the booksellers after tho publishing films have pio ducad them. We believe that the ait of bread making a? thus perfected will eventually supersede the old hand and foot labour upon the dough, and the old piactice of decomposing the bread itself to procuie the gas that is to lighten it But we do not believe that the change will be prejudicial to the interests of a trade that in its piessnt state seems to be one of the least profitable and the most unhealthy that a man can follow. —Abndaed from \\\"Alt the Year Hound \\\"Death of Mr. Awued Bonn.—With a deep regiet, that will be bhaied by all who knew him intimately, and with the consciousness of a loss that will be widely felt through .all literary and diamatic circles, we have to announce the sudden decease of Mr Alfied Bunn, who expired on Thursday, of apoplexy. For some time past Mr. Bunn had letiied fiooi public life , and about two yeard ago he fixed hid residence at Boulogne, wheie Ins death took place. Theie weio no indications of such an abrupt teunination of his life being near ; indeed, though a twelvemonth since seriously indisposed, his health and spirits had lately so much impioved that the very article which appears in our papei this week, \\\" A Summary View of a Wintery Year,\\\" had pioceeded from his everactive pen but a few horns bofoic Latter! 7hs had embiaced the Roman Catholic fa'th, and to the leligious offices of that creed he paid strict attention. In theatucal mattou he continued, however, to feel a lively interest ; and to this paper he was a frequent and a valued contributor, especially as the writer if tiose articles from \\\" Our own Boulogne Correspondent.\\\" Mr. Bunn's association with thestage dated fiom n very early period. In 1826, he was a manager of the Birmingham Theatre ; and, after succeeding (Juptaiu Polhill in the lesseeship of Drurylane and Oovcnt-gardrn Theatres, conducting both through two remaikably consecutive seasons, which will be especially lemembeied for the strong companies engaged at eaoh house, find tho prosperous career of the Opeia of \\\"Oustavus the Third,\\\" that was brought out by him at the latter From 1835 until 1848 he conti oiled tho destinies of Drury-lane Theatre, giving the public the popular series of Opeiaa by Balfe, tho advantage of the engagement of all the most attractive Opawtic, Choiogiaphic and Diamatic stars that could bo obtained, amongst whom must be mentioned the names of Madame Malibian, Duveinay, Forrest, dialled Kean, and Macready, and, in the face of great difficulties, at last maintaining tho attractions of the establishment at high pi ices, whilst the rates of admission were being «ene ally reduced elsewheie Besides being known as the author of the lihrtlo of the \\\" Bohemiau Girl,\\\" Mr. Alfred Bunn will be remembered as the writer of thioe volumes called \\\"The Stage, Before and Behind tho Cm tarn ;\\\" and a brochure, railed \\\" A Woid with Punch,\\\" both of which publications cieated a great sensation at the time A sketch of the vailed and interesting incidents of his life we shall endeavour to piesent to our leaders next week? and in the interval would only ask them to lumember that the position which Mr. Alfied Bunn occupied for so long a period bcfoie the public was one that could not fail to affect a variety of conflicting interests, and that much of the animosity which he had to endure aiose from the difficulties which ha had to encounter, in attempting, at tho same time, to satisfy both the piofession and the public In pitvate life Mr. Bunn was a warm fuend and a highly intelligent and most agreeable companion — EmFoul ign Tradf Mahics. — In the House of Lords on the 12th Felnumy the Loid Chancellor laid on the table ft 811 which he said was of the highest importance to manufacturers., protecting them not from fair competition but from fiaud. Its chief object wan to protect them horn the grievances under which they suffered fioni the forging of tiade maiks. The greit manufactiuers all had a par ticular mark to denote their manufactures, which was known at home and abroad, and which was a guaiante3 of the genuineness and excellence of the article, and it had of late yeaia becomo e\\\\tieuiely common to forge tV mark of a successful, eminent, and celebrated manufactiuei, thus not only depriving him of the piofits to which he was justly entitled, butbiiugingchsgiace on him in addition, because the articles thus imposed on the public with false maiks were geneially of an inferior quality and of a fraudulent nature. In most othei couutues the forging of such marks was a cume punishable by law, but with us the only remedy was to apply to tha Court of Chancery for an injunction and to bring an action at law for damages. This hid been found to be a dilatory, expensive, and uncertain lemedy. The noble and learned lord on the bench near him (Lord Chelmsford) would remember that one of the last cases on which he was engaged at the close of brilliant caieei at the bar was an action biought by Farina, of Cologne, against some person who had forged his mark on the Eau de Cologne. There was a long litigation, both in the Court of Chanceiy and m the Queen's Bench, and a very imperfect remedy was obtained by the person who had been wionged. It was proposed by tln« Bill that the foiging of these maiks, 01 the sale of ai tides beii ing forged maiks, with the intent to dcfiaud, should be made a misdemeanour, punishable by fine and linpnsonment This it \\\\\\\\ as to be hoped w ould pi ovide an effectual lemedy. Another benefit would be confer led on raanufActuieis in this way. Our countrymen Buffeied very much by these piactices in foreign countries, and when they complained the answer they leceived was, \\\" You do nothing to piotect us in our country. Theie must be reciprocity. If you would give us a lemedy we would most villingly listen to your complaint \\\" When this Bill beeune law that leproach would cease, and our manufacturer, it was to be hoped, would have ] ustice done to them 111 all jiai ts of the world There was also a fraudulent piaotice of using maiks to denote the quantity and quality of goods which were utteily false , foi instuiee, a milk would be put on a piece of to fabnc to denote that it was of a certain length, when it was in fact much shortei, or of a ceitain nntcunlt lie piopo&ed to make a fraud of this soit a misdemeanour , puiiislublc m the same way. It had aKo, o; Lite, become an exceedingly common practice, in tire & lie of w oiks of aii, to put on them the name, maik, 01 lnonogiam of some paiticulai aitist, so as to deceive even counouseuib Under this Bill, if the name or mouogiam of any particular aitist weie fraudulently attached to any woik of .vit, so as, to deceive , peisuns reasonably skilled i>i such aiticles, that would be a misdemeanour , punishable by fine and nnpiibonI ment.The English as Rum us — At a lailwuy banquet at Rajmahal (Bengal), Loid Canning, who piesided, said — \\\" Cicntlenien, — -It is no use to deny or conceal i*-, foi it is Known to all the woild — we, Englishmen, with all our great national charactenstics, are not, as a pec- | pie, conciliatoiy 01 atti active God forbid that any of ! vlb should fetl .ashamed of his national character, 01 wish it to be othei than it i». But none amongst us will deny that the veiy virtues of that chaiacter aie not seldom cxaggeiated into faults We aie powerful in body and in mind, and aie pioud of that power We are selfreliant, and justly so, — and we like to show our selfreliance We aie conscious of oui high put poses and enlightenment, and we are apt to look down upon those whose motives we believe to bi lebs woithy than our own, oi whom we icgard as debased in ignoiance — and we do not caie to conceal oiu feelings. These feelings are not inconsistent with out national gieatne33. In the days of slaveiy, Englishmen weie amongst the • haidest taskmasteis that the Afiican ever had ; but England did not hesitate to spend her gold and her blood lavishly foi the suppiessiou of the blave trade, and we pouied out oiu £20,000,000 like water, when we found that it was the only means by which to lid ourselves of the curse of slaveiy But gentlemen, no people, whatever then condition, will patiently beai to be heated by their rulers as though they weie much less than men ; less capable of light feeling than those who mle them. If we attempt, individually or collectively, to do this — if we neglect to win the heaits of those over whom Providence has placed us — if, instead of seeking to inspiie them with confidence, we take for our maxim that the people of India should be governed as a conqueied people, which, as I inideistand it, means that they should be governed by sheer force ; if, in our piide or impatience we letuse to show foibeaiance and indulgence to the weakness and shoitcoimngs which offend us, we shall not woitlnly lcpiesent England in the gieat woik which lies befoiehor, and we shall absiuedly fail to accomplish it 1 hay c been led to make these observat'ons because I desire to express the gieat satisfaction with which I have v\\\\ atched the spirit in which all dealings or relations of the East India Railway Company have been conducted. It beais upon a subject the impoitance of which I cannot exaggeiate , and in fegaid to which there is not one Englishman m India, fiom the Governor-gen-i eial in his camp to the aitizan seeking his> livelihood, who does n t cai ry in his keeping a poi tion of the credit and good name of England.PnoTOCrUAi'HS n\\\\ Machineim . — A highly interesting and singulai paper was lead befoie the American Pho tographic Society lately, and is reported in the \\\" Architect's Journal\\\" of New Yoik. By this paper it appears that 12,000 photographs or steieographs ,in houi can now be produced from a single negative by means of condensed ox focalised light and simple machinery by a ciank. i A sheet of ozdmary paper, sensitised, was exhibited, containing 300 of those photogiaphs. Mr. Chas Fontayne, of Cincinnati, Ohio, is the inventor of the piocess. The piepaied or sensitised paper is simply pasbed in a continuous sheet, before the negative, in a box, wheie condensed light is made to penetiato through the negative and imprebs its image upon the papei, which it does in 03 of a second for each impiession. The condensing lens is 7 inches in diameter. Thus, as it is said, '• the illustrations for a book, having all the exquisite beauty and perfection of the photogiaph, may bo turned out by the vwe of this machine with a rapidity wholly undieamt of either in plain printing or in lithography.\\\" The cost of engiavmg 1 , alio, will of couise bo dispensed with All toits of diawingq, too, may bo thus multiplied, as well as actual objects photographed or steieogiaphed, in cheap and endlesi profusion. The AicJutecii' Join nul publishes a print thus pioducodfrom a lough bketch by the oidinaiy ammomo mtiate process.Holloway's Ointment and Pills. — Erysipelas, is one of the most unsightly anrl alarming of the diseases of the skin, it pioceeds like otheis fiom an mi puie state of the blood and fluids. If Holloway's Ointment and Pills be used, it is most astonishing the immediate and wonderful effect they will have upon this disouler ; in fact they may be said to act like a charm, destroying the virus, and acting as a depurifyer of the blood, preventing other diseases from entering the system, <vnd restoring the suffeier to sound and robust health after .all other means have failed.Cjbkwihes or thb Amkuican Cbnsus — The present census haa revealed some items of interest in the internal history of the country. In \\\"Upper Aroostooks, Ma, there is a family of twenty-seven children, all living at homo, and the mother but forty years of age. And m New Milford, Conn., are two maideu ladies who aie each two yoara younger than than they -\\\\veie in 1820! Productiveness ok Bees. —Some lnvei of bees lately taken by Mr. \\\"W. Wilson have yielded so miich as 951bs. each of honey comb. Sixty pouuds in Eng land is considered a remarkable yield. — New Zealand Bptcttttor.\"\n"
     ]
    }
   ],
   "source": [
    "%%bash -s $line\n",
    "echo \"print the $1 document:\"\n",
    "sed -n \"$1p\" ../data/dataset/sample/train/train.csv"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Recommendation 2:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# linux line number is based on 0\n",
    "line = list_sim[1] +1 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "print the 2724006 document:\n",
      "28113158\tUNKNOWN\t\"UNKNOWN![| %ypvii grcnte,ey I '-lie purpose .of iablishment of ■''\\\"'twas? - '*v*S r -»» * J \\\"|i\\\"*'V<(tri'J- i 'i is -'-• 'ti — | ior>|fceiv and if ''Si.'applied to tWial Govemm |S|jfor,stesii%ng S^^^^fk&r.,,\"\n"
     ]
    }
   ],
   "source": [
    "%%bash -s $line\n",
    "echo \"print the $1 document:\"\n",
    "sed -n \"$1p\" ../data/dataset/sample/train/train.csv"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Recommendation 3:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# linux line number is based on 0\n",
    "line = list_sim[2] +1 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "print the 1142501 document:\n",
      "12564083\tTHEIR WAY.\t\"THEIR WAY.Bi ffi h ' : \\\" How de toifrtit-yitofrt\\\"Nothing.\\\". , : , \\\" Then why did she go ahoppinfr P\\\" \\\"To see if she could find anything that would make her want Bourething.\\\"\"\n"
     ]
    }
   ],
   "source": [
    "%%bash -s $line\n",
    "echo \"print the $1 document:\"\n",
    "sed -n \"$1p\" ../data/dataset/sample/train/train.csv"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Recommendation 4:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "# linux line number is based on 0\n",
    "line = list_sim[3] +1 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "print the 8685 document:\n",
      "1929443\tCORRESPONDENCE\tCORRESPONDENCETl)0 Kilitov.is fk'sivoiw of alTonlijij* every vciis.»»:il)lo futility fi'v the f'iswissiun of' iu'W.ic sjiil'jccfs,; ljut it must lie (iistinct y ini(lcr.stno(! liiat lie is in no way rc.si»)iisiliie for the opinions oxjiriwbvil. by (jonvsj.on-,, 'ilcnts. ■ ' ■ . ' - ■ .\n"
     ]
    }
   ],
   "source": [
    "%%bash -s $line\n",
    "echo \"print the $1 document:\"\n",
    "sed -n \"$1p\" ../data/dataset/sample/train/train.csv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "# linux line number is based on 0\n",
    "line = list_sim[4] +1 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "print the 2839840 document:\n",
      "29000415\tUntitled\t\"Tlte e.itii'e iloor of the liirfrc carpet eliow ~„,„„ N,;-l and Cloho mado >mn a bio for JL.jldir\\\" their nsii.-il ChiistinaH cshtbii.on Th-'.-'IiDW in consequence complete, it contains v wun.lurful n.iscelb.ne.us •icv'.'-tm-nt of unpfnl u.-fii'lcx, in addition to toys awl tiich other things to please and gratify the vigitpV; Inaction invited. }Vdiuis;;i'ju tret-'v ' ;\"\n"
     ]
    }
   ],
   "source": [
    "%%bash -s $line\n",
    "echo \"print the $1 document:\"\n",
    "sed -n \"$1p\" ../data/dataset/sample/train/train.csv"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**We see the results seem make no sense. The process should be fine, the reason is that the doc-topic matrix of the test corpus is not reasonable, which means MALLET did not give a acceptable inferring output. We take the further digging about MALLET inferring issues as future work.**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "31"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sc.stop()\n",
    "gc.collect()"
   ]
  }
 ],
 "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}