{ "metadata": { "name": "", "signature": "sha256:cf6a6e8c91c04ea1dd3ba6c9c15b50053c3018a599c4c01519bb0cd2ed1df8d5" }, "nbformat": 3, "nbformat_minor": 0, "worksheets": [ { "cells": [ { "cell_type": "heading", "level": 2, "metadata": {}, "source": [ "Topic visualizations of a Poisson-Gamma DEF (size 50-25-10) trained on 1,000 wikipedia articles" ] }, { "cell_type": "code", "collapsed": false, "input": [ "%pylab inline\n", "%load_ext autoreload\n", "%autoreload 2\n", "\n", "import pandas\n", "import sys" ], "language": "python", "metadata": {}, "outputs": [ { "output_type": "stream", "stream": "stdout", "text": [ "Populating the interactive namespace from numpy and matplotlib\n", "The autoreload extension is already loaded. To reload it, use:\n", " %reload_ext autoreload\n" ] }, { "output_type": "stream", "stream": "stderr", "text": [ "WARNING: pylab import has clobbered these variables: ['box', 'linalg', 'text', 'random', 'power', 'info', 'fft']\n", "`%matplotlib` prevents importing * from pylab and numpy\n" ] } ], "prompt_number": 17 }, { "cell_type": "code", "collapsed": false, "input": [ "# specific imports\n", "sys.path += ['../scripts/']\n", "from utils import *\n", "from pyx import *\n", "from wand.image import Image as WImage\n", "from def_visualization import *" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 18 }, { "cell_type": "code", "collapsed": false, "input": [ "word_list = read_words('./vocab.dat')\n", "experiment_dir = '../experiments/def_wikipedia_1434725288667'" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 19 }, { "cell_type": "code", "collapsed": false, "input": [ "t = map(softrect, load_bin_model(experiment_dir + '/train_iter01000.model.bin'))" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 20 }, { "cell_type": "code", "collapsed": false, "input": [ "W0_shape, W0_scale, z0_shape, z0_scale = t[:4] \n", "z1_shape, z1_scale, z2_shape, z2_scale = t[4:8] \n", "W1_shape, W1_scale, W2_shape, W2_scale = t[8:]\n", "W0_mean = W0_shape * W0_scale\n", "W1_mean = W1_shape * W1_scale\n", "W2_mean = W2_shape * W2_scale" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 21 }, { "cell_type": "heading", "level": 4, "metadata": {}, "source": [ "First layer topics" ] }, { "cell_type": "code", "collapsed": false, "input": [ "g = top_words(W0_mean, word_list, k=10, W_shape=W0_shape, show_weight=True)" ], "language": "python", "metadata": {}, "outputs": [ { "output_type": "stream", "stream": "stdout", "text": [ "Topic 0\n", "state 0.341688850243 17.8120743811\n", "party 0.266396299919 18.664387271\n", "national 0.125530215608 14.7028305909\n", "alabama 0.11259996231 18.566584038\n", "election 0.109002015749 18.6986190985\n", "governor 0.103520611551 19.1626345865\n", "arkansas 0.102276140417 17.2617603328\n", "vote 0.0972261548324 19.3580183463\n", "government 0.0851871717519 13.6195711971\n", "democratic 0.0843305840673 17.8221570027\n", "\n", "\n", "Topic 1\n", "novel 0.0874780204024 17.974768033\n", "play 0.0719818291177 17.2042960767\n", "stories 0.0567413823244 18.6402343842\n", "best 0.0504194294446 13.31251659\n", "murder 0.0501055232976 18.1011363256\n", "short 0.0496254920266 16.1028957148\n", "story 0.0493768255188 16.301524876\n", "book 0.0490051249294 11.4179205446\n", "character 0.043847161499 14.1439288862\n", "plays 0.0430505500842 17.9414175636\n", "\n", "\n", "Topic 2\n", "species 0.15095027799 19.3060348502\n", "family 0.116784969268 17.0666613815\n", "plants 0.076895045062 19.2684370357\n", "food 0.063013448874 16.8469749536\n", "animals 0.060496255285 17.2980767712\n", "genus 0.0544655345468 18.6177252814\n", "order 0.0489690231599 16.0086328966\n", "plant 0.0463831916658 17.4085005782\n", "agricultural 0.0399812151507 16.8478295401\n", "fish 0.0373797518917 15.8324416873\n", "\n", "\n", "Topic 3\n", "law 0.171070885209 17.2186974745\n", "property 0.0978889222137 17.5503117987\n", "state 0.0970791637472 14.0629411442\n", "act 0.0954956425588 14.9691121185\n", "free 0.0867581420438 15.7583295903\n", "society 0.0825887476592 15.3661049488\n", "austrian 0.0812394962357 16.6194508596\n", "rights 0.0798770794722 16.116761491\n", "private 0.0769261658317 17.4075753599\n", "market 0.073828080647 16.4319282312\n", "\n", "\n", "Topic 4\n", "apple 0.096268917424 21.4406549343\n", "system 0.0539576895435 18.9358058188\n", "software 0.0443668415609 21.3435837218\n", "computer 0.0437460107971 18.6823876947\n", "company 0.0264559503121 17.6486107991\n", "systems 0.0264456014955 16.8511851479\n", "new 0.0260737423571 13.6918737347\n", "released 0.0256231883601 19.8171289049\n", "based 0.0229260300411 15.9004979595\n", "memory 0.0205695643695 19.1369236001\n", "\n", "\n", "Topic 5\n", "chinese 2.28572962341 16.7068081329\n", "dishes 1.40568747796 17.4765079005\n", "popular 1.07993196725 13.261924076\n", "meat 1.04522945734 16.6940978868\n", "american 1.0191087221 0.711096701148\n", "called 1.01869937294 7.92132223452\n", "native 1.00442584534 13.4047564375\n", "made 0.941376863797 12.4831203534\n", "italian 0.90869609306 13.5612268907\n", "region 0.839748034905 11.4228004228\n", "\n", "\n", "Topic 6\n", "center 4.89074700042 18.081899016\n", "style 4.78943606662 17.7665207871\n", "background 3.2751123702 18.5698847037\n", "color 0.800568681039 16.8402558401\n", "text 0.680361580371 14.3542492806\n", "open 0.300284514558 16.6334682359\n", "hard 0.276584486409 17.0426369046\n", "year 0.144556078053 11.4404098586\n", "anna 0.129620736267 15.2657720938\n", "final 0.126181435731 13.5476914595\n", "\n", "\n", "Topic 7\n", "house 9.56848857185 10.285130994\n", "abbey 7.67401940855 13.8953137911\n", "monks 6.84056726282 13.3628780796\n", "side 5.54007994627 12.1932866584\n", "buildings 5.23463087207 10.9583074393\n", "superior 4.21698851135 11.9323853379\n", "church 4.03003449422 6.19925295354\n", "wall 3.95261049535 12.7084904901\n", "order 3.63334013408 7.53281407306\n", "rule 3.42773454182 10.3708694318\n", "\n", "\n", "Topic 8\n", "medicine 0.10461629492 18.9609286397\n", "treatment 0.0674128982554 20.4322557225\n", "medical 0.0634410254566 19.3880635873\n", "effects 0.0587334983134 19.446963451\n", "health 0.0564545599254 19.1908781292\n", "patients 0.0534872220053 20.2797777891\n", "alternative 0.0525662305252 17.4907750313\n", "pain 0.0427614050201 18.6135697133\n", "disease 0.0418795443011 20.0688088526\n", "evidence 0.0369882759094 17.4566770277\n", "\n", "\n", "Topic 9\n", "sea 0.0511888838158 19.8053886477\n", "south 0.0394126333324 16.965886564\n", "north 0.0372162120427 16.9874166924\n", "island 0.033497524383 18.4792174506\n", "east 0.02840929755 16.9578726509\n", "area 0.0280653717397 15.984714159\n", "river 0.0273337659915 17.7098019301\n", "islands 0.0272092796387 19.8389363107\n", "right 0.0265155605595 14.2117490501\n", "west 0.022486853032 15.8070987633\n", "\n", "\n", "Topic 10\n", "court 0.963023110649 16.2954840821\n", "car 0.691062653122 17.0344315083\n", "martin 0.593550319247 15.8937160862\n", "cars 0.580702651332 16.957520749\n", "age 0.389623619814 13.9831993752\n", "appeal 0.384827115083 16.1530862625\n", "engine 0.357712152625 14.1196017854\n", "assault 0.354880316129 15.830561311\n", "consent 0.292085801717 15.5747847808\n", "ford 0.280037685301 15.9573217147\n", "\n", "\n", "Topic 11\n", "bell 67.2249819335 19.7122349258\n", "att 27.5842215992 16.2649254748\n", "telephone 25.7783017078 17.5841179943\n", "sound 10.9405587401 13.4103126774\n", "deaf 9.78499640147 16.3894898948\n", "patent 8.69505534019 15.8362952125\n", "company 8.43536407139 10.319681472\n", "speech 7.68915017519 13.214257024\n", "graham 6.97884423909 14.7042350049\n", "alexander 5.42671091883 7.11303420372\n", "\n", "\n", "Topic 12\n", "art 1.95779707296 18.8280887107\n", "artist 0.320479822378 17.8155014918\n", "works 0.286406227871 15.6192527256\n", "andy 0.220495844489 17.2458515858\n", "arts 0.208083702799 16.1239890116\n", "artists 0.205436956558 17.2591797303\n", "work 0.175649933572 12.9761323147\n", "style 0.162774020606 7.69639290706\n", "artistic 0.161267757443 17.137899335\n", "new 0.155753663372 8.04085946539\n", "\n", "\n", "Topic 13\n", "city 0.144144515344 19.304849193\n", "university 0.0641667040477 17.3228845224\n", "built 0.035784503859 17.6104735359\n", "school 0.0331530324802 16.8699483108\n", "building 0.0297461440961 18.588368701\n", "college 0.0279653989094 17.1323214157\n", "area 0.0255879472083 15.0074680202\n", "park 0.0254137309206 18.9965515109\n", "house 0.022589672836 16.2924686729\n", "town 0.0205728039754 17.2511470313\n", "\n", "\n", "Topic 14\n", "apollo 0.553905643217 20.0911535157\n", "mission 0.368909508991 20.2556548265\n", "crew 0.32937903968 20.0780997831\n", "moon 0.275414892987 19.8117652688\n", "space 0.22663756336 18.6659476196\n", "earth 0.203303962422 18.1626753727\n", "first 0.191101079383 12.7649776789\n", "flight 0.165406761416 18.3970661225\n", "command 0.146940859324 19.3243997029\n", "landing 0.130577077039 18.4214463325\n", "\n", "\n", "Topic 15\n", "chain 80.0469519269 19.7247096298\n", "chains 38.7583338038 17.0614451365\n", "stock 24.654604166 15.8706686353\n", "game 18.6868495655 9.3780303507\n", "exchange 16.1857521344 10.8142287175\n", "shares 14.871109103 14.5301998113\n", "three 13.3816919543 6.97078718171\n", "board 10.7039148349 10.648490399\n", "acquired 10.4554895833 12.2231308629\n", "share 8.57640281327 11.246750981\n", "\n", "\n", "Topic 16\n", "language 0.0985669880624 19.1221728017\n", "arabic 0.0676594165731 20.8226760833\n", "languages 0.0512059415407 19.6573191004\n", "line 0.0377916095158 18.0626128631\n", "written 0.0303443578806 17.0129804529\n", "letters 0.0297546760325 18.903266464\n", "century 0.0282500096754 13.3801234447\n", "word 0.0279372514171 15.5021414248\n", "height 0.027340401861 17.995801398\n", "modern 0.0267237675307 15.3009484281\n", "\n", "\n", "Topic 17\n", "architecture 14.7584655888 15.1140072528\n", "social 11.5715325478 15.9842974614\n", "design 11.0518982154 13.631152098\n", "game 8.74101489115 14.7243246107\n", "base 7.77498631438 15.0792962684\n", "experience 6.83270411054 14.2047230367\n", "society 6.72911377555 11.2510244306\n", "requirements 5.62119747107 15.0277855193\n", "projects 4.39860444579 11.9167936809\n", "civilization 4.11042955869 13.2431967135\n", "\n", "\n", "Topic 18\n", "war 0.366636753142 13.4511219213\n", "air 0.257807875216 13.8546875614\n", "armour 0.247328141589 16.4528091975\n", "first 0.199427200368 7.0498551\n", "navy 0.172113903412 13.9443833331\n", "gun 0.154344302785 15.9295838725\n", "military 0.143964861247 13.6221699072\n", "flight 0.1414632128 9.73264468788\n", "world 0.136606940285 10.5126647941\n", "deck 0.110985815805 13.4369026211\n", "\n", "\n", "Topic 19\n", "african 30.7918911167 16.951177008\n", "americans 19.8098324199 16.6420315273\n", "american 16.8632911127 8.25570047146\n", "black 14.6287790373 15.0747227904\n", "day 9.52279511114 9.15004143247\n", "states 8.75055571676 8.04532498206\n", "tree 5.69955210284 13.4709155869\n", "trees 4.47181008592 13.0004534916\n", "native 3.99982056215 9.10649548528\n", "groups 3.99569617458 8.8733248122\n", "\n", "\n", "Topic 20\n", "empire 0.0191191638089 18.7594512713\n", "army 0.017535017032 18.7884392105\n", "roman 0.0155115170914 17.9714932066\n", "augustus 0.0128666162309 18.5791811493\n", "emperor 0.0111348218527 18.0194199484\n", "battle 0.0106147466746 18.2623251084\n", "rome 0.00961715107612 18.8251213556\n", "death 0.00948030023419 13.1677544926\n", "war 0.0085714085162 12.7165143612\n", "military 0.0085543337529 15.9501600782\n", "\n", "\n", "Topic 21\n", "ancient 0.0942549470866 18.0241689329\n", "greek 0.0779371001225 16.0785599947\n", "century 0.0629646126596 12.1568551499\n", "athens 0.0549776897891 14.8974609755\n", "period 0.0481868627021 15.5047475697\n", "known 0.0458343706066 9.53914049012\n", "egypt 0.0449946088346 17.2780135921\n", "temple 0.0387892587739 16.6752958805\n", "made 0.035772857763 9.99106734142\n", "stone 0.0347791626673 16.9030432292\n", "\n", "\n", "Topic 22\n", "title 60.3018572433 16.7566126515\n", "page 48.5331739981 16.729271373\n", "comment 39.6126931255 18.0061301284\n", "animals 26.9062085576 11.6192957456\n", "preserve 26.4730190096 15.4390849065\n", "animal 22.8141844215 12.2333475307\n", "farm 21.4926249487 13.9034036971\n", "book 19.8458888251 7.82061495477\n", "space 18.9529627135 11.2448011398\n", "napoleon 18.3287604385 13.6739327378\n", "\n", "\n", "Topic 23" ] }, { "output_type": "stream", "stream": "stdout", "text": [ "\n", "english 0.352921361714 17.3925807727\n", "american 0.186004499778 8.91551906388\n", "british 0.152514889194 15.8247183377\n", "word 0.148355517692 15.8880516679\n", "united 0.131854254036 11.3459728743\n", "words 0.111496852015 15.7834468518\n", "term 0.0872322438853 14.3140830299\n", "case 0.0846720569044 13.1642314485\n", "states 0.0839148280437 10.5229448002\n", "common 0.0815505919748 14.0988946982\n", "\n", "\n", "Topic 24\n", "computer 0.103151330019 16.2151434089\n", "time 0.0967879389877 14.5826065986\n", "ascii 0.0783216787987 18.4526213229\n", "code 0.0679424215628 17.8547020999\n", "standard 0.0661618948283 16.8236924034\n", "characters 0.0631680426962 17.1903835705\n", "language 0.0623123236943 13.8003098047\n", "data 0.0549262249758 17.0555616484\n", "program 0.0516359578167 16.4303673147\n", "character 0.0456923761187 15.9735493375\n", "\n", "\n", "Topic 25\n", "star 0.409317041294 16.7760624726\n", "angle 0.380780536417 15.1331340136\n", "stars 0.362429638917 17.7659766024\n", "earth 0.344122643185 14.7419895859\n", "sun 0.324056014183 16.5276860698\n", "angles 0.240742962285 14.2616154232\n", "system 0.230185624397 13.2142779772\n", "light 0.229713565894 14.2527948403\n", "objects 0.210094852684 15.9728530241\n", "moon 0.196964538909 14.6553371604\n", "\n", "\n", "Topic 26\n", "american 0.176619587151 22.5274110852\n", "actor 0.0354383553142 22.8294379272\n", "english 0.0311531756093 21.4713598991\n", "french 0.0307929554825 21.4143067511\n", "british 0.023944673033 20.4284299514\n", "canadian 0.0182633652119 23.057861751\n", "german 0.0164184728337 20.596811523\n", "first 0.0153804006432 17.3160659847\n", "john 0.0146162015882 20.8642546769\n", "writer 0.0143316819279 22.0777229493\n", "\n", "\n", "Topic 27\n", "series 1.2658579276 14.9871941536\n", "present 0.656620984896 15.8589672916\n", "characters 0.644922922094 13.4610739876\n", "character 0.571789866977 13.9334146341\n", "show 0.458730452062 14.148140221\n", "animated 0.453775793989 16.7971477637\n", "adventures 0.353784169476 17.2079775704\n", "children 0.287266467946 10.8407996078\n", "alien 0.276550968453 14.5241954566\n", "game 0.269674556165 11.8574704753\n", "\n", "\n", "Topic 28\n", "women 43.8255203957 17.1284385208\n", "hill 21.4630786556 9.4840992476\n", "sex 18.6032543966 14.4685718025\n", "woman 18.4697184626 14.2316162614\n", "thomas 15.2182308074 6.97418405575\n", "female 12.9991487246 12.1448211816\n", "sexual 12.1546643631 11.7682620325\n", "breast 11.2551533788 12.8439041745\n", "university 8.35750054531 2.06404900532\n", "life 8.32496477182 7.72438967437\n", "\n", "\n", "Topic 29\n", "jews 4.32531232017 18.9894360611\n", "jewish 2.74859763552 18.7899712674\n", "camp 1.00744795236 16.7677025324\n", "arab 0.991049112241 16.7453466883\n", "prisoners 0.545711522813 16.4275234484\n", "jesus 0.442340117047 13.2869465929\n", "christians 0.420565433314 14.3012782331\n", "lewis 0.40227720475 15.9692581852\n", "german 0.39200102675 10.6804391949\n", "israel 0.389619298715 14.2414601641\n", "\n", "\n", "Topic 30\n", "left 74.6406083355 18.983248148\n", "right 56.7124726017 17.7169093969\n", "partial 31.1695517029 19.2442232877\n", "time 10.155313071 6.49910360375\n", "sum 9.53443120998 14.9694417489\n", "tilde 9.22202908644 17.4796508498\n", "balance 8.42310800757 16.1698538715\n", "tree 5.68040030223 3.44349959451\n", "text 5.57478686536 7.40012573536\n", "factor 5.51849960988 7.69035388465\n", "\n", "\n", "Topic 31\n", "philosophy 0.0158643440518 21.2117845589\n", "book 0.0114641964102 16.202735982\n", "human 0.0111977413822 16.9639459788\n", "life 0.00995174929771 13.8608653004\n", "world 0.00953893660735 12.3184555056\n", "smith 0.0092771026091 17.96096267\n", "works 0.00849844937101 16.4942302781\n", "acts 0.0081756094542 17.2585490412\n", "nature 0.0078970343741 17.8217691524\n", "god 0.00779859640033 13.3256274153\n", "\n", "\n", "Topic 32\n", "team 0.0775154895153 21.5209363691\n", "first 0.0708712628727 16.3871648636\n", "league 0.0650324087714 21.563885793\n", "season 0.0543639171829 21.9282208898\n", "club 0.0464696440211 21.0589824667\n", "two 0.0380232377926 14.4442718589\n", "won 0.0350129523348 19.9664529543\n", "time 0.0348088432074 12.6432040817\n", "game 0.0344853408617 19.0403600588\n", "games 0.033911037349 20.351750164\n", "\n", "\n", "Topic 33\n", "section 24.5077708929 15.7650700899\n", "steel 20.0466239713 15.5058827221\n", "francisco 12.219222291 14.6507402209\n", "appears 10.2476938181 12.4191241533\n", "sections 7.83892645969 13.9558871419\n", "mentioned 7.68136226304 13.1924396433\n", "railroad 7.04735488924 12.6530815064\n", "james 6.57674763457 10.8921601481\n", "line 5.31201451445 4.36184630954\n", "john 5.06238646094 6.74978969636\n", "\n", "\n", "Topic 34\n", "god 0.0591746171486 16.9821495312\n", "name 0.0465587969857 14.9256558474\n", "son 0.0366289521902 15.7488008725\n", "king 0.0327199633067 12.3908843594\n", "abraham 0.0327072358572 17.6040799226\n", "greek 0.0269851282799 12.4140300384\n", "father 0.0241489910543 13.4846270521\n", "people 0.0231901522555 12.3988648977\n", "two 0.0206976159736 8.9580807908\n", "time 0.0183506796495 8.25993942315\n", "\n", "\n", "Topic 35\n", "german 8.43451012224 14.2695741177\n", "austria 7.21274601841 13.4695276119\n", "van 6.04762647303 16.4639287484\n", "dutch 5.10433832486 18.4016257609\n", "austrian 4.26134262589 12.3442520096\n", "germany 3.28539773859 14.0191288087\n", "der 2.47823833975 13.0482776572\n", "netherlands 2.06949812731 15.2370601171\n", "european 1.99575638233 11.1597099467\n", "vienna 1.52173432399 11.6352532829\n", "\n", "\n", "Topic 36\n", "number 0.0546221657408 15.0457494619\n", "numbers 0.0510620264647 18.0584401414\n", "theory 0.0507540298845 15.5505295174\n", "value 0.0451672165378 17.6531491807\n", "function 0.0406285856729 17.473932509\n", "choice 0.0377026511465 17.1047495891\n", "two 0.0328251888954 11.087050694\n", "real 0.0327650950389 17.3109564155\n", "argument 0.0324739451825 18.1356528746\n", "called 0.0305952102826 11.6873971094\n", "\n", "\n", "Topic 37\n", "first 0.0174563445405 11.4946712919\n", "years 0.0139636512965 14.4619447154\n", "two 0.0121624166391 13.1949320478\n", "life 0.0115214357371 15.1445206588\n", "time 0.0108724257154 12.4620270325\n", "new 0.00942431517629 10.7339289128\n", "year 0.00938387984911 13.274527602\n", "made 0.00757165311399 12.008319745\n", "death 0.00712545172439 13.9935728946\n", "early 0.00634405517136 12.3725231069\n", "\n", "\n", "Topic 38\n", "bones 12.4564539903 10.9223796094\n", "skull 11.4925210601 10.5718172004\n", "large 11.128403527 5.72804336819\n", "remains 10.0199979789 9.67898509478\n", "teeth 8.34404439705 6.86933282408\n", "formation 6.72887620545 5.94375360813\n", "prey 6.63073702395 5.74588926243\n", "specimen 6.59688354258 7.24031774647\n", "like 5.99789671802 3.76349717343\n", "genus 5.96688888018 3.31118158019\n", "\n", "\n", "Topic 39\n", "music 0.244972826584 17.8342655012\n", "song 0.190309371999 19.0118789653\n", "band 0.0944033119767 17.4848839644\n", "released 0.0845995696729 16.3712266614\n", "musical 0.0820588942433 17.9900230919\n", "group 0.0761197425987 14.4743393707\n", "single 0.0648762989933 14.7062683368\n", "opera 0.0595234626515 16.8929554716\n", "instruments 0.0557158792845 16.0886470388\n", "songs 0.055653065197 17.4987168846\n", "\n", "\n", "Topic 40\n", "work 0.0293017900623 15.6008489627\n", "university 0.0290519795207 16.3456440709\n", "theory 0.0200509888907 16.1529795799\n", "history 0.019062125521 14.4447236763\n", "analysis 0.0153338474385 18.0215874271\n", "science 0.0152223618461 17.9572037633\n", "research 0.0150101142178 17.4580956559\n", "study 0.0144442620431 17.2798292849\n", "century 0.0139684709618 11.2465966788\n", "modern 0.013704719145 12.7701234199\n", "\n", "\n", "Topic 41\n", "alexander 0.0891172615464 20.6880766512\n", "king 0.050938546488 18.2440157818\n", "iii 0.0269262334183 20.2059536881\n", "father 0.0220617372982 17.310019078\n", "albert 0.0211780891724 18.6096382093\n", "son 0.0189550070147 16.0749147113\n", "married 0.0179399093892 18.6900729414\n", "mother 0.0174602409039 17.8245379121\n", "pope 0.0170940077016 17.6526791033\n", "emperor 0.0158657764548 16.0223909121\n", "\n", "\n", "Topic 42\n", "lincoln 1.34329646738 19.6429201467\n", "war 0.970581891594 16.0646583506\n", "union 0.890187442627 18.7283198349\n", "states 0.870160074086 16.6179901548\n", "jackson 0.771964858748 18.221523822\n", "johnson 0.645375656891 17.2070004835\n", "president 0.489057365177 17.5835452908\n", "south 0.483327599115 14.7463713883\n", "confederate 0.446940925265 19.6415794098\n", "battle 0.431240314986 16.0787582294\n", "\n", "\n", "Topic 43\n", "japanese 4.03295370172 12.0436303784\n", "bow 3.12859375651 16.6983652287\n", "academy 3.03927816998 14.2943321325\n", "japan 2.83589522093 12.3102419673\n", "arrow 2.66658296934 16.8919110348\n", "arrows 2.60334983619 16.6170822966\n", "motion 2.28804982758 13.4093070332\n", "shooting 2.24050098067 15.4805244475\n", "picture 2.06731877132 15.0098915218\n", "shoot 1.99589094574 16.0449339455\n", "\n", "\n", "Topic 44\n", "british 1.35350309005 17.6798067962\n", "american 0.755891177695 10.1905496888\n", "new 0.717735123159 13.6793880354\n", "war 0.613531735752 14.1001481531\n", "congress 0.485162695929 16.5986862113\n", "states 0.443072970912 11.1869240374\n", "united 0.419213596292 9.53066193622\n", "colonies 0.404503184825 16.8722704904\n", "americans 0.398173971214 15.839108381\n", "washington 0.358830671017 15.7295751313\n", "\n", "\n", "Topic 45\n", "united 0.0309884068905 14.9163922905\n", "states 0.0243862033573 12.6939782214\n", "national 0.0190522966395 14.2428402357\n", "world 0.0181891829145 11.5932667532\n", "new 0.0178009724864 10.5687878887\n", "government 0.0177616412483 14.4764729212\n", "international 0.0173453478923 16.0116601792\n", "million 0.0144795400216 13.8790944906\n", "first 0.0140622426405 6.38302411075\n", "president 0.0124958481946 15.561041063\n", "\n", "\n", "Topic 46\n", "england 2.6975768559 17.0423911477\n", "alfred 1.92078374037 17.0395137256\n", "ashes 1.2636940781 13.6353682659\n", "english 1.20478510204 12.656038024\n", "saxon 1.11993914519 17.3754012034\n", "australia 1.11208906996 13.9044711631\n", "series 1.1038239231 10.1592295859\n", "made 0.874513292089 8.43425558681\n", "test 0.780996622269 13.7767475789\n", "history 0.383214486659 9.27136452185\n", "\n", "\n", "Topic 47" ] }, { "output_type": "stream", "stream": "stdout", "text": [ "\n", "population 0.0683053848068 19.3791388812\n", "state 0.0426021965616 10.1117065907\n", "country 0.0347772879553 16.6607371818\n", "government 0.0291466818687 15.0863457983\n", "oil 0.0289120230656 19.2602132577\n", "austin 0.0282136508976 17.1273919957\n", "people 0.0280655328187 12.2603344851\n", "area 0.0269760162347 14.2724050542\n", "largest 0.0258339969143 17.0792173207\n", "north 0.0251832471447 12.1863605135\n", "\n", "\n", "Topic 48\n", "church 0.236262214986 20.1054647124\n", "catholic 0.083546948926 20.1540154076\n", "churches 0.0692832120781 19.91681444\n", "roman 0.0553210784849 17.8036947583\n", "christian 0.0540839167125 18.4209030155\n", "god 0.0519898916139 15.6859775669\n", "christ 0.0502033508683 19.2768535134\n", "communion 0.0475419465384 19.4339406221\n", "jesus 0.0418865741648 17.8034227535\n", "succession 0.0406334056411 17.9630261897\n", "\n", "\n", "Topic 49\n", "two 0.00790115444999 15.9613621892\n", "form 0.00629226553919 18.8151466182\n", "called 0.00586023420865 16.883415402\n", "known 0.00522969723338 14.7108355963\n", "acid 0.00502750640659 22.3280698112\n", "different 0.00491934891282 18.5603299608\n", "number 0.00483822526832 16.2455163946\n", "water 0.00430048223097 20.3760500394\n", "first 0.00390821538419 11.1863832478\n", "found 0.00382394929034 15.3755557562\n", "\n", "\n" ] } ], "prompt_number": 22 }, { "cell_type": "heading", "level": 4, "metadata": {}, "source": [ "Second layer groups" ] }, { "cell_type": "code", "collapsed": false, "input": [ "#For each group we show: \n", "# 1) Most probable words\n", "# 2) Top topics per group including their probability\n", "# 3) The top word per topic\n", "g = top_groups(W1_mean, W0_mean, word_list, k1=3, k=10, show_weight=True)" ], "language": "python", "metadata": {}, "outputs": [ { "output_type": "stream", "stream": "stdout", "text": [ "group 0\n", "first work time new two american years university bell known\n", "37 40 49\n", "0.123751054626 0.0670412148524 0.0492089040261\n", "37 first years two life time new year made death early\n", "40 work university theory history analysis science research study century modern\n", "49 two form called known acid different number water first found\n", "group 1\n", "language arabic languages english line word century written form modern\n", "16 49 31\n", "0.292228244563 0.106261696067 0.0218289318463\n", "16 language arabic languages line written letters century word height modern\n", "49 two form called known acid different number water first found\n", "31 philosophy book human life world smith works acts nature god\n", "group 2\n", "city sea south north area east first island river world\n", "9 45 13\n", "0.310430258223 0.152166165333 0.111145216503\n", "09 sea south north island east area river islands right west\n", "45 united states national world new government international million first president\n", "13 city university built school building college area park house town\n", "group 3\n", "earth apollo first moon two star mission system angle stars\n", "49 37 9\n", "0.497952527712 0.0597400613132 0.0323294311621\n", "49 two form called known acid different number water first found\n", "37 first years two life time new year made death early\n", "09 sea south north island east area river islands right west\n", "group 4\n", "american english united first known new chinese called made states\n", "49 26 45\n", "0.071235625642 0.0283982709684 0.027125217071\n", "49 two form called known acid different number water first found\n", "26 american actor english french british canadian german first john writer\n", "45 united states national world new government international million first president" ] }, { "output_type": "stream", "stream": "stdout", "text": [ "\n", "group 5\n", "state states united national government american first party new war\n", "45 20 37\n", "0.230613331935 0.0763507056892 0.0717367931338\n", "45 united states national world new government international million first president\n", "20 empire army roman augustus emperor battle rome death war military\n", "37 first years two life time new year made death early\n", "group 6\n", "art first new life time two years series book made\n", "37 31 1\n", "0.459670164508 0.107075141686 0.0552817397358\n", "37 first years two life time new year made death early\n", "31 philosophy book human life world smith works acts nature god\n", "01 novel play stories best murder short story book character plays\n", "group 7\n", "number two called form numbers theory value different group function\n", "49 36 31\n", "1.12950134718 0.176591138862 0.0304059421632\n", "49 two form called known acid different number water first found\n", "36 number numbers theory value function choice two real argument called\n", "31 philosophy book human life world smith works acts nature god\n", "group 8\n", "title page comment animals preserve book animal first text farm\n", "40 41 31\n", "0.019120022751 0.01819631244 0.0155918749974\n", "40 work university theory history analysis science research study century modern\n", "41 alexander king iii father albert son married mother pope emperor\n", "31 philosophy book human life world smith works acts nature god\n", "group 9\n", "alexander king son father death time first name emperor iii\n", "20 41 37\n", "0.62544525721 0.514171626621 0.457520129121\n", "20 empire army roman augustus emperor battle rome death war military\n", "41 alexander king iii father albert son married mother pope emperor\n", "37 first years two life time new year made death early\n", "group 10\n", "title page first jews comment american art bell book animals\n", "40 49 20\n", "0.00725809447236 0.00651797387484 0.00533356093004\n", "40 work university theory history analysis science research study century modern\n", "49 two form called known acid different number water first found\n", "20 empire army roman augustus emperor battle rome death war military" ] }, { "output_type": "stream", "stream": "stdout", "text": [ "\n", "group 11\n", "right style center sea south first north left two background\n", "49 9 13\n", "0.0350008855936 0.0316154390833 0.0051195824249\n", "49 two form called known acid different number water first found\n", "09 sea south north island east area river islands right west\n", "13 city university built school building college area park house town\n", "group 12\n", "music song first new band american released group musical single\n", "37 39 49\n", "0.0626980805129 0.0312935523508 0.0204838021343\n", "37 first years two life time new year made death early\n", "39 music song band released musical group single opera instruments songs\n", "49 two form called known acid different number water first found\n", "group 13\n", "two known species form called found family different acid number\n", "49 2 8\n", "2.22501014371 0.10618909788 0.0803082897559\n", "49 two form called known acid different number water first found\n", "02 species family plants food animals genus order plant agricultural fish\n", "08 medicine treatment medical effects health patients alternative pain disease evidence\n", "group 14\n", "war first air armour navy world style gun flight center\n", "49 18 37\n", "0.216780631851 0.0170477108923 0.0160443884176\n", "49 two form called known acid different number water first found\n", "18 war air armour first navy gun military flight world deck\n", "37 first years two life time new year made death early\n", "group 15\n", "american title page first university english british comment war world\n", "40 31 9\n", "0.022935360636 0.0157983083931 0.0139405899314\n", "40 work university theory history analysis science research study century modern\n", "31 philosophy book human life world smith works acts nature god\n", "09 sea south north island east area river islands right west\n", "group 16\n", "apple computer system time first two called software systems based\n", "49 4 24\n", "0.897405858146 0.209953984253 0.0884871372906\n", "49 two form called known acid different number water first found\n", "04 apple system software computer company systems new released based memory\n", "24 computer time ascii code standard characters language data program character\n", "group 17" ] }, { "output_type": "stream", "stream": "stdout", "text": [ "\n", "american actor english french british first canadian german war john\n", "26 32 20\n", "1.14965776722 0.112563199814 0.0664066982786\n", "26 american actor english french british canadian german first john writer\n", "32 team first league season club two won time game games\n", "20 empire army roman augustus emperor battle rome death war military\n", "group 18\n", "center style art title american background page first comment work\n", "40 20 45\n", "0.0199687944108 0.0082935653619 0.00714065787009\n", "40 work university theory history analysis science research study century modern\n", "20 empire army roman augustus emperor battle rome death war military\n", "45 united states national world new government international million first president\n", "group 19\n", "american bell first title english left british time series right\n", "40 26 41\n", "0.00782300082724 0.00686779520655 0.00576757519762\n", "40 work university theory history analysis science research study century modern\n", "26 american actor english french british canadian german first john writer\n", "41 alexander king iii father albert son married mother pope emperor\n", "group 20\n", "american city first art new united world time university english\n", "45 13 26\n", "0.0101304475376 0.00818056384726 0.00679153636349\n", "45 united states national world new government international million first president\n", "13 city university built school building college area park house town\n", "26 american actor english french british canadian german first john writer\n", "group 21\n", "center style background color text open city hard first year\n", "37 41 13\n", "0.0508642793604 0.0136565638516 0.0136385090268\n", "37 first years two life time new year made death early\n", "41 alexander king iii father albert son married mother pope emperor\n", "13 city university built school building college area park house town\n", "group 22\n", "first work church university time century theory life world history" ] }, { "output_type": "stream", "stream": "stdout", "text": [ "\n", "37 31 40\n", "0.667715206946 0.640392266466 0.630158854901\n", "37 first years two life time new year made death early\n", "31 philosophy book human life world smith works acts nature god\n", "40 work university theory history analysis science research study century modern\n", "group 23\n", "american first city known series state university time new system\n", "49 40 37\n", "0.0167104439144 0.0103562024122 0.0101832173823\n", "49 two form called known acid different number water first found\n", "40 work university theory history analysis science research study century modern\n", "37 first years two life time new year made death early\n", "group 24\n", "title page first comment german art english american space states\n", "37 49 31\n", "0.0110106096513 0.0072003622595 0.00380316872361\n", "37 first years two life time new year made death early\n", "49 two form called known acid different number water first found\n", "31 philosophy book human life world smith works acts nature god\n" ] } ], "prompt_number": 23 }, { "cell_type": "heading", "level": 4, "metadata": {}, "source": [ "Third layer super groups" ] }, { "cell_type": "code", "collapsed": false, "input": [ "g = top_supers(W2_mean, W1_mean, W0_mean, word_list, k2=5, k1=5, k=6, show_weight=True)" ], "language": "python", "metadata": {}, "outputs": [ { "output_type": "stream", "stream": "stdout", "text": [ "SUPER 0\n", "american first english french actor two\n", "17 9 8 13 4\n", "4.75704920931 3.15565077209 2.1770083874 1.93258185117 1.88924506013\n", "group 0\n", "american actor english french british first\n", "26 32 20 45 13\n", "1.14965776722 0.112563199814 0.0664066982786 0.0284960443054 0.0251595129677\n", "26 american actor english french british canadian\n", "32 team first league season club two\n", "20 empire army roman augustus emperor battle\n", "45 united states national world new government\n", "13 city university built school building college\n", "group 1\n", "alexander king son father death time\n", "20 41 37 34 21\n", "0.62544525721 0.514171626621 0.457520129121 0.15864688743 0.0801005782384\n", "20 empire army roman augustus emperor battle\n", "41 alexander king iii father albert son\n", "37 first years two life time new\n", "34 god name son king abraham greek\n", "21 ancient greek century athens period known\n", "group 2\n", "title page comment animals preserve book\n", "40 41 31 9 37\n", "0.019120022751 0.01819631244 0.0155918749974 0.0150055669722 0.0117947661997\n", "40 work university theory history analysis science\n", "41 alexander king iii father albert son\n", "31 philosophy book human life world smith\n", "09 sea south north island east area\n", "37 first years two life time new\n", "group 3\n", "two known species form called found\n", "49 2 8 40 9\n", "2.22501014371 0.10618909788 0.0803082897559 0.0301053167546 0.0241809472234\n", "49 two form called known acid different" ] }, { "output_type": "stream", "stream": "stdout", "text": [ "\n", "02 species family plants food animals genus\n", "08 medicine treatment medical effects health patients\n", "40 work university theory history analysis science\n", "09 sea south north island east area\n", "group 4\n", "american english united first known new\n", "49 26 45 41 16\n", "0.071235625642 0.0283982709684 0.027125217071 0.0136646046541 0.0081071754926\n", "49 two form called known acid different\n", "26 american actor english french british canadian\n", "45 united states national world new government\n", "41 alexander king iii father albert son\n", "16 language arabic languages line written letters\n", "\n", "\n", "\n", "SUPER 1\n", "american first alexander king time two\n", "9 22 17 24 15\n", "6.49425227545 3.32671751736 2.73831780234 2.45471784812 2.39315655145\n", "group 0\n", "alexander king son father death time\n", "20 41 37 34 21\n", "0.62544525721 0.514171626621 0.457520129121 0.15864688743 0.0801005782384\n", "20 empire army roman augustus emperor battle\n", "41 alexander king iii father albert son\n", "37 first years two life time new\n", "34 god name son king abraham greek\n", "21 ancient greek century athens period known\n", "group 1\n", "first work church university time century\n", "37 31 40 48 49\n", "0.667715206946 0.640392266466 0.630158854901 0.102368555247 0.063817550481\n", "37 first years two life time new\n", "31 philosophy book human life world smith\n", "40 work university theory history analysis science\n", "48 church catholic churches roman christian god\n", "49 two form called known acid different\n", "group 2\n", "american actor english french british first" ] }, { "output_type": "stream", "stream": "stdout", "text": [ "\n", "26 32 20 45 13\n", "1.14965776722 0.112563199814 0.0664066982786 0.0284960443054 0.0251595129677\n", "26 american actor english french british canadian\n", "32 team first league season club two\n", "20 empire army roman augustus emperor battle\n", "45 united states national world new government\n", "13 city university built school building college\n", "group 3\n", "title page first comment german art\n", "37 49 31 45 48\n", "0.0110106096513 0.0072003622595 0.00380316872361 0.00376626028959 0.00315199944027\n", "37 first years two life time new\n", "49 two form called known acid different\n", "31 philosophy book human life world smith\n", "45 united states national world new government\n", "48 church catholic churches roman christian god\n", "group 4\n", "american title page first university english\n", "40 31 9 26 13\n", "0.022935360636 0.0157983083931 0.0139405899314 0.00748074473526 0.00383432643002\n", "40 work university theory history analysis science\n", "31 philosophy book human life world smith\n", "09 sea south north island east area\n", "26 american actor english french british canadian\n", "13 city university built school building college\n", "\n", "\n", "\n", "SUPER 2\n", "american first two time new known\n", "4 6 1 20 15\n", "2.8263173065 1.89090698541 1.79355751268 1.65932237772 1.45175531048\n", "group 0\n", "american english united first known new\n", "49 26 45 41 16\n", "0.071235625642 0.0283982709684 0.027125217071 0.0136646046541 0.0081071754926\n", "49 two form called known acid different\n", "26 american actor english french british canadian\n", "45 united states national world new government\n", "41 alexander king iii father albert son\n", "16 language arabic languages line written letters" ] }, { "output_type": "stream", "stream": "stdout", "text": [ "\n", "group 1\n", "art first new life time two\n", "37 31 1 26 39\n", "0.459670164508 0.107075141686 0.0552817397358 0.0205030646673 0.0151238162613\n", "37 first years two life time new\n", "31 philosophy book human life world smith\n", "01 novel play stories best murder short\n", "26 american actor english french british canadian\n", "39 music song band released musical group\n", "group 2\n", "language arabic languages english line word\n", "16 49 31 21 23\n", "0.292228244563 0.106261696067 0.0218289318463 0.0192625218568 0.0178765402334\n", "16 language arabic languages line written letters\n", "49 two form called known acid different\n", "31 philosophy book human life world smith\n", "21 ancient greek century athens period known\n", "23 english american british word united words\n", "group 3\n", "american city first art new united\n", "45 13 26 31 2\n", "0.0101304475376 0.00818056384726 0.00679153636349 0.00656465394158 0.00419999877609\n", "45 united states national world new government\n", "13 city university built school building college\n", "26 american actor english french british canadian\n", "31 philosophy book human life world smith\n", "02 species family plants food animals genus\n", "group 4\n", "american title page first university english\n", "40 31 9 26 13\n", "0.022935360636 0.0157983083931 0.0139405899314 0.00748074473526 0.00383432643002\n", "40 work university theory history analysis science\n", "31 philosophy book human life world smith\n", "09 sea south north island east area\n", "26 american actor english french british canadian\n", "13 city university built school building college" ] }, { "output_type": "stream", "stream": "stdout", "text": [ "\n", "\n", "\n", "\n", "SUPER 3\n", "american first two known time new\n", "13 0 2 9 3\n", "2.82243386268 2.21896672798 2.04509879731 1.95592322442 1.95016416095\n", "group 0\n", "two known species form called found\n", "49 2 8 40 9\n", "2.22501014371 0.10618909788 0.0803082897559 0.0301053167546 0.0241809472234\n", "49 two form called known acid different\n", "02 species family plants food animals genus\n", "08 medicine treatment medical effects health patients\n", "40 work university theory history analysis science\n", "09 sea south north island east area\n", "group 1\n", "first work time new two american\n", "37 40 49 41 26\n", "0.123751054626 0.0670412148524 0.0492089040261 0.00821391196035 0.00792996231651\n", "37 first years two life time new\n", "40 work university theory history analysis science\n", "49 two form called known acid different\n", "41 alexander king iii father albert son\n", "26 american actor english french british canadian\n", "group 2\n", "city sea south north area east\n", "9 45 13 49 37\n", "0.310430258223 0.152166165333 0.111145216503 0.104547062836 0.103731187619\n", "09 sea south north island east area\n", "45 united states national world new government\n", "13 city university built school building college\n", "49 two form called known acid different\n", "37 first years two life time new\n", "group 3\n", "alexander king son father death time\n", "20 41 37 34 21\n", "0.62544525721 0.514171626621 0.457520129121 0.15864688743 0.0801005782384\n", "20 empire army roman augustus emperor battle\n", "41 alexander king iii father albert son\n", "37 first years two life time new\n", "34 god name son king abraham greek" ] }, { "output_type": "stream", "stream": "stdout", "text": [ "\n", "21 ancient greek century athens period known\n", "group 4\n", "earth apollo first moon two star\n", "49 37 9 45 14\n", "0.497952527712 0.0597400613132 0.0323294311621 0.031345394341 0.0176203386152\n", "49 two form called known acid different\n", "37 first years two life time new\n", "09 sea south north island east area\n", "45 united states national world new government\n", "14 apollo mission crew moon space earth\n", "\n", "\n", "\n", "SUPER 4\n", "american first two known time new\n", "11 13 24 14 0\n", "2.98318536855 2.90556098163 2.86472288123 2.76474675494 2.29943545596\n", "group 0\n", "right style center sea south first\n", "49 9 13 16 8\n", "0.0350008855936 0.0316154390833 0.0051195824249 0.00460929709985 0.00413551495019\n", "49 two form called known acid different\n", "09 sea south north island east area\n", "13 city university built school building college\n", "16 language arabic languages line written letters\n", "08 medicine treatment medical effects health patients\n", "group 1\n", "two known species form called found\n", "49 2 8 40 9\n", "2.22501014371 0.10618909788 0.0803082897559 0.0301053167546 0.0241809472234\n", "49 two form called known acid different\n", "02 species family plants food animals genus\n", "08 medicine treatment medical effects health patients\n", "40 work university theory history analysis science\n", "09 sea south north island east area" ] }, { "output_type": "stream", "stream": "stdout", "text": [ "\n", "group 2\n", "title page first comment german art\n", "37 49 31 45 48\n", "0.0110106096513 0.0072003622595 0.00380316872361 0.00376626028959 0.00315199944027\n", "37 first years two life time new\n", "49 two form called known acid different\n", "31 philosophy book human life world smith\n", "45 united states national world new government\n", "48 church catholic churches roman christian god\n", "group 3\n", "war first air armour navy world\n", "49 18 37 4 24\n", "0.216780631851 0.0170477108923 0.0160443884176 0.0139357212822 0.00580174531109\n", "49 two form called known acid different\n", "18 war air armour first navy gun\n", "37 first years two life time new\n", "04 apple system software computer company systems\n", "24 computer time ascii code standard characters\n", "group 4\n", "first work time new two american\n", "37 40 49 41 26\n", "0.123751054626 0.0670412148524 0.0492089040261 0.00821391196035 0.00792996231651\n", "37 first years two life time new\n", "40 work university theory history analysis science\n", "49 two form called known acid different\n", "41 alexander king iii father albert son\n", "26 american actor english french british canadian" ] }, { "output_type": "stream", "stream": "stdout", "text": [ "\n", "\n", "\n", "\n", "SUPER 5\n", "american first two time alexander known\n", "9 13 22 2 7\n", "4.05535769811 3.00469723187 2.45702684166 2.00469571026 1.83722897902\n", "group 0\n", "alexander king son father death time\n", "20 41 37 34 21\n", "0.62544525721 0.514171626621 0.457520129121 0.15864688743 0.0801005782384\n", "20 empire army roman augustus emperor battle\n", "41 alexander king iii father albert son\n", "37 first years two life time new\n", "34 god name son king abraham greek\n", "21 ancient greek century athens period known\n", "group 1\n", "two known species form called found\n", "49 2 8 40 9\n", "2.22501014371 0.10618909788 0.0803082897559 0.0301053167546 0.0241809472234\n", "49 two form called known acid different\n", "02 species family plants food animals genus\n", "08 medicine treatment medical effects health patients\n", "40 work university theory history analysis science\n", "09 sea south north island east area\n", "group 2\n", "first work church university time century\n", "37 31 40 48 49\n", "0.667715206946 0.640392266466 0.630158854901 0.102368555247 0.063817550481\n", "37 first years two life time new\n", "31 philosophy book human life world smith\n", "40 work university theory history analysis science\n", "48 church catholic churches roman christian god\n", "49 two form called known acid different\n", "group 3\n", "city sea south north area east\n", "9 45 13 49 37\n", "0.310430258223 0.152166165333 0.111145216503 0.104547062836 0.103731187619\n", "09 sea south north island east area\n", "45 united states national world new government\n", "13 city university built school building college\n", "49 two form called known acid different" ] }, { "output_type": "stream", "stream": "stdout", "text": [ "\n", "37 first years two life time new\n", "group 4\n", "number two called form numbers theory\n", "49 36 31 40 41\n", "1.12950134718 0.176591138862 0.0304059421632 0.0284431497353 0.00634382429922\n", "49 two form called known acid different\n", "36 number numbers theory value function choice\n", "31 philosophy book human life world smith\n", "40 work university theory history analysis science\n", "41 alexander king iii father albert son\n", "\n", "\n", "\n", "SUPER 6\n", "first american two time known called\n", "13 9 7 22 3\n", "3.23476421683 3.19948173599 2.87804491507 2.71757206235 2.12204748022\n", "group 0\n", "two known species form called found\n", "49 2 8 40 9\n", "2.22501014371 0.10618909788 0.0803082897559 0.0301053167546 0.0241809472234\n", "49 two form called known acid different\n", "02 species family plants food animals genus\n", "08 medicine treatment medical effects health patients\n", "40 work university theory history analysis science\n", "09 sea south north island east area\n", "group 1\n", "alexander king son father death time\n", "20 41 37 34 21\n", "0.62544525721 0.514171626621 0.457520129121 0.15864688743 0.0801005782384\n", "20 empire army roman augustus emperor battle\n", "41 alexander king iii father albert son\n", "37 first years two life time new\n", "34 god name son king abraham greek\n", "21 ancient greek century athens period known\n", "group 2\n", "number two called form numbers theory\n", "49 36 31 40 41\n", "1.12950134718 0.176591138862 0.0304059421632 0.0284431497353 0.00634382429922\n", "49 two form called known acid different\n", "36 number numbers theory value function choice" ] }, { "output_type": "stream", "stream": "stdout", "text": [ "\n", "31 philosophy book human life world smith\n", "40 work university theory history analysis science\n", "41 alexander king iii father albert son\n", "group 3\n", "first work church university time century\n", "37 31 40 48 49\n", "0.667715206946 0.640392266466 0.630158854901 0.102368555247 0.063817550481\n", "37 first years two life time new\n", "31 philosophy book human life world smith\n", "40 work university theory history analysis science\n", "48 church catholic churches roman christian god\n", "49 two form called known acid different\n", "group 4\n", "earth apollo first moon two star\n", "49 37 9 45 14\n", "0.497952527712 0.0597400613132 0.0323294311621 0.031345394341 0.0176203386152\n", "49 two form called known acid different\n", "37 first years two life time new\n", "09 sea south north island east area\n", "45 united states national world new government\n", "14 apollo mission crew moon space earth\n", "\n", "\n", "\n", "SUPER 7\n", "american first new world city state\n", "5 2 22 17 21\n", "27.9054541582 27.7047876212 8.27769805108 5.23249340745 4.14337040695\n", "group 0\n", "state states united national government american\n", "45 20 37 49 47\n", "0.230613331935 0.0763507056892 0.0717367931338 0.060455248185 0.0317483782998\n", "45 united states national world new government\n", "20 empire army roman augustus emperor battle\n", "37 first years two life time new\n", "49 two form called known acid different\n", "47 population state country government oil austin\n", "group 1\n", "city sea south north area east\n", "9 45 13 49 37" ] }, { "output_type": "stream", "stream": "stdout", "text": [ "\n", "0.310430258223 0.152166165333 0.111145216503 0.104547062836 0.103731187619\n", "09 sea south north island east area\n", "45 united states national world new government\n", "13 city university built school building college\n", "49 two form called known acid different\n", "37 first years two life time new\n", "group 2\n", "first work church university time century\n", "37 31 40 48 49\n", "0.667715206946 0.640392266466 0.630158854901 0.102368555247 0.063817550481\n", "37 first years two life time new\n", "31 philosophy book human life world smith\n", "40 work university theory history analysis science\n", "48 church catholic churches roman christian god\n", "49 two form called known acid different\n", "group 3\n", "american actor english french british first\n", "26 32 20 45 13\n", "1.14965776722 0.112563199814 0.0664066982786 0.0284960443054 0.0251595129677\n", "26 american actor english french british canadian\n", "32 team first league season club two\n", "20 empire army roman augustus emperor battle\n", "45 united states national world new government\n", "13 city university built school building college\n", "group 4\n", "center style background color text open\n", "37 41 13 16 6\n", "0.0508642793604 0.0136565638516 0.0136385090268 0.00982958806594 0.00878233259808\n", "37 first years two life time new\n", "41 alexander king iii father albert son\n", "13 city university built school building college\n", "16 language arabic languages line written letters\n", "06 center style background color text open\n", "\n", "\n", "\n", "SUPER 8\n", "american first time two system apple\n", "16 1 13 24 20\n", "12.947402875 2.59675688685 2.56259511361 2.53526605861 2.23985033591\n", "group 0\n", "apple computer system time first two" ] }, { "output_type": "stream", "stream": "stdout", "text": [ "\n", "49 4 24 45 36\n", "0.897405858146 0.209953984253 0.0884871372906 0.0669129535887 0.0483910176081\n", "49 two form called known acid different\n", "04 apple system software computer company systems\n", "24 computer time ascii code standard characters\n", "45 united states national world new government\n", "36 number numbers theory value function choice\n", "group 1\n", "language arabic languages english line word\n", "16 49 31 21 23\n", "0.292228244563 0.106261696067 0.0218289318463 0.0192625218568 0.0178765402334\n", "16 language arabic languages line written letters\n", "49 two form called known acid different\n", "31 philosophy book human life world smith\n", "21 ancient greek century athens period known\n", "23 english american british word united words\n", "group 2\n", "two known species form called found\n", "49 2 8 40 9\n", "2.22501014371 0.10618909788 0.0803082897559 0.0301053167546 0.0241809472234\n", "49 two form called known acid different\n", "02 species family plants food animals genus\n", "08 medicine treatment medical effects health patients\n", "40 work university theory history analysis science\n", "09 sea south north island east area\n", "group 3\n", "title page first comment german art\n", "37 49 31 45 48\n", "0.0110106096513 0.0072003622595 0.00380316872361 0.00376626028959 0.00315199944027\n", "37 first years two life time new\n", "49 two form called known acid different\n", "31 philosophy book human life world smith\n", "45 united states national world new government\n", "48 church catholic churches roman christian god\n", "group 4\n", "american city first art new united\n", "45 13 26 31 2\n", "0.0101304475376 0.00818056384726 0.00679153636349 0.00656465394158 0.00419999877609\n", "45 united states national world new government\n", "13 city university built school building college\n", "26 american actor english french british canadian\n", "31 philosophy book human life world smith\n", "02 species family plants food animals genus" ] }, { "output_type": "stream", "stream": "stdout", "text": [ "\n", "\n", "\n", "\n", "SUPER 9\n", "first american time two new years\n", "6 9 3 22 8\n", "16.8411061162 6.00162928157 5.79468595137 4.98180637303 2.69681724246\n", "group 0\n", "art first new life time two\n", "37 31 1 26 39\n", "0.459670164508 0.107075141686 0.0552817397358 0.0205030646673 0.0151238162613\n", "37 first years two life time new\n", "31 philosophy book human life world smith\n", "01 novel play stories best murder short\n", "26 american actor english french british canadian\n", "39 music song band released musical group\n", "group 1\n", "alexander king son father death time\n", "20 41 37 34 21\n", "0.62544525721 0.514171626621 0.457520129121 0.15864688743 0.0801005782384\n", "20 empire army roman augustus emperor battle\n", "41 alexander king iii father albert son\n", "37 first years two life time new\n", "34 god name son king abraham greek\n", "21 ancient greek century athens period known\n", "group 2\n", "earth apollo first moon two star\n", "49 37 9 45 14\n", "0.497952527712 0.0597400613132 0.0323294311621 0.031345394341 0.0176203386152\n", "49 two form called known acid different\n", "37 first years two life time new\n", "09 sea south north island east area\n", "45 united states national world new government\n", "14 apollo mission crew moon space earth\n", "group 3\n", "first work church university time century\n", "37 31 40 48 49\n", "0.667715206946 0.640392266466 0.630158854901 0.102368555247 0.063817550481\n", "37 first years two life time new\n", "31 philosophy book human life world smith\n", "40 work university theory history analysis science\n", "48 church catholic churches roman christian god" ] }, { "output_type": "stream", "stream": "stdout", "text": [ "\n", "49 two form called known acid different\n", "group 4\n", "title page comment animals preserve book\n", "40 41 31 9 37\n", "0.019120022751 0.01819631244 0.0155918749974 0.0150055669722 0.0117947661997\n", "40 work university theory history analysis science\n", "41 alexander king iii father albert son\n", "31 philosophy book human life world smith\n", "09 sea south north island east area\n", "37 first years two life time new\n", "\n", "\n", "\n" ] } ], "prompt_number": 24 } ], "metadata": {} } ] }