| <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> |
| <html xmlns="http://www.w3.org/1999/xhtml"> |
| <head> |
| <meta http-equiv="Content-Type" content="text/xhtml;charset=UTF-8"/> |
| <meta http-equiv="X-UA-Compatible" content="IE=9"/> |
| <meta name="generator" content="Doxygen 1.8.4"/> |
| <title>MADlib: lda.sql_in Source File</title> |
| <link href="tabs.css" rel="stylesheet" type="text/css"/> |
| <script type="text/javascript" src="jquery.js"></script> |
| <script type="text/javascript" src="dynsections.js"></script> |
| <link href="navtree.css" rel="stylesheet" type="text/css"/> |
| <script type="text/javascript" src="resize.js"></script> |
| <script type="text/javascript" src="navtree.js"></script> |
| <script type="text/javascript"> |
| $(document).ready(initResizable); |
| $(window).load(resizeHeight); |
| </script> |
| <link href="search/search.css" rel="stylesheet" type="text/css"/> |
| <script type="text/javascript" src="search/search.js"></script> |
| <script type="text/javascript"> |
| $(document).ready(function() { searchBox.OnSelectItem(0); }); |
| </script> |
| <script type="text/x-mathjax-config"> |
| MathJax.Hub.Config({ |
| extensions: ["tex2jax.js", "TeX/AMSmath.js", "TeX/AMSsymbols.js"], |
| jax: ["input/TeX","output/HTML-CSS"], |
| }); |
| </script><script src="../mathjax/MathJax.js"></script> |
| <link href="doxygen.css" rel="stylesheet" type="text/css" /> |
| </head> |
| <body> |
| <div id="top"><!-- do not remove this div, it is closed by doxygen! --> |
| <div id="titlearea"> |
| <table cellspacing="0" cellpadding="0"> |
| <tbody> |
| <tr style="height: 56px;"> |
| <td style="padding-left: 0.5em;"> |
| <div id="projectname">MADlib |
|  <span id="projectnumber">1.0</span> <span style="font-size:10pt; font-style:italic"><a href="../latest/./lda_8sql__in_source.html"> A newer version is available</a></span> |
| </div> |
| <div id="projectbrief">User Documentation</div> |
| </td> |
| </tr> |
| </tbody> |
| </table> |
| </div> |
| <!-- end header part --> |
| <!-- Generated by Doxygen 1.8.4 --> |
| <script type="text/javascript"> |
| var searchBox = new SearchBox("searchBox", "search",false,'Search'); |
| </script> |
| <div id="navrow1" class="tabs"> |
| <ul class="tablist"> |
| <li><a href="index.html"><span>Main Page</span></a></li> |
| <li><a href="modules.html"><span>Modules</span></a></li> |
| <li> |
| <div id="MSearchBox" class="MSearchBoxInactive"> |
| <span class="left"> |
| <img id="MSearchSelect" src="search/mag_sel.png" |
| onmouseover="return searchBox.OnSearchSelectShow()" |
| onmouseout="return searchBox.OnSearchSelectHide()" |
| alt=""/> |
| <input type="text" id="MSearchField" value="Search" accesskey="S" |
| onfocus="searchBox.OnSearchFieldFocus(true)" |
| onblur="searchBox.OnSearchFieldFocus(false)" |
| onkeyup="searchBox.OnSearchFieldChange(event)"/> |
| </span><span class="right"> |
| <a id="MSearchClose" href="javascript:searchBox.CloseResultsWindow()"><img id="MSearchCloseImg" border="0" src="search/close.png" alt=""/></a> |
| </span> |
| </div> |
| </li> |
| </ul> |
| </div> |
| </div><!-- top --> |
| <div id="side-nav" class="ui-resizable side-nav-resizable"> |
| <div id="nav-tree"> |
| <div id="nav-tree-contents"> |
| <div id="nav-sync" class="sync"></div> |
| </div> |
| </div> |
| <div id="splitbar" style="-moz-user-select:none;" |
| class="ui-resizable-handle"> |
| </div> |
| </div> |
| <script type="text/javascript"> |
| $(document).ready(function(){initNavTree('lda_8sql__in_source.html','');}); |
| </script> |
| <div id="doc-content"> |
| <!-- window showing the filter options --> |
| <div id="MSearchSelectWindow" |
| onmouseover="return searchBox.OnSearchSelectShow()" |
| onmouseout="return searchBox.OnSearchSelectHide()" |
| onkeydown="return searchBox.OnSearchSelectKey(event)"> |
| <a class="SelectItem" href="javascript:void(0)" onclick="searchBox.OnSelectItem(0)"><span class="SelectionMark"> </span>All</a><a class="SelectItem" href="javascript:void(0)" onclick="searchBox.OnSelectItem(1)"><span class="SelectionMark"> </span>Files</a><a class="SelectItem" href="javascript:void(0)" onclick="searchBox.OnSelectItem(2)"><span class="SelectionMark"> </span>Functions</a><a class="SelectItem" href="javascript:void(0)" onclick="searchBox.OnSelectItem(3)"><span class="SelectionMark"> </span>Groups</a></div> |
| |
| <!-- iframe showing the search results (closed by default) --> |
| <div id="MSearchResultsWindow"> |
| <iframe src="javascript:void(0)" frameborder="0" |
| name="MSearchResults" id="MSearchResults"> |
| </iframe> |
| </div> |
| |
| <div class="header"> |
| <div class="headertitle"> |
| <div class="title">lda.sql_in</div> </div> |
| </div><!--header--> |
| <div class="contents"> |
| <a href="lda_8sql__in.html">Go to the documentation of this file.</a><div class="fragment"><div class="line"><a name="l00001"></a><span class="lineno"> 1</span> <span class="comment">/* ----------------------------------------------------------------------- */</span><span class="comment">/** </span></div> |
| <div class="line"><a name="l00002"></a><span class="lineno"> 2</span> <span class="comment"> *</span></div> |
| <div class="line"><a name="l00003"></a><span class="lineno"> 3</span> <span class="comment"> * @file lda.sql_in</span></div> |
| <div class="line"><a name="l00004"></a><span class="lineno"> 4</span> <span class="comment"> *</span></div> |
| <div class="line"><a name="l00005"></a><span class="lineno"> 5</span> <span class="comment"> * @brief SQL functions for Latent Dirichlet Allocation</span></div> |
| <div class="line"><a name="l00006"></a><span class="lineno"> 6</span> <span class="comment"> * @date Dec 2012</span></div> |
| <div class="line"><a name="l00007"></a><span class="lineno"> 7</span> <span class="comment"> *</span></div> |
| <div class="line"><a name="l00008"></a><span class="lineno"> 8</span> <span class="comment"> * @sa For an introduction to Latent Dirichlet Allocation models, see the </span></div> |
| <div class="line"><a name="l00009"></a><span class="lineno"> 9</span> <span class="comment"> module description \ref grp_lda.</span></div> |
| <div class="line"><a name="l00010"></a><span class="lineno"> 10</span> <span class="comment"> *</span></div> |
| <div class="line"><a name="l00011"></a><span class="lineno"> 11</span> <span class="comment"> */</span><span class="comment">/* ------------------------------------------------------------------------*/</span></div> |
| <div class="line"><a name="l00012"></a><span class="lineno"> 12</span> </div> |
| <div class="line"><a name="l00013"></a><span class="lineno"> 13</span> m4_include(`SQLCommon.m4<span class="stringliteral">')</span></div> |
| <div class="line"><a name="l00014"></a><span class="lineno"> 14</span> <span class="stringliteral"></span><span class="comment"></span></div> |
| <div class="line"><a name="l00015"></a><span class="lineno"> 15</span> <span class="comment">/**</span></div> |
| <div class="line"><a name="l00016"></a><span class="lineno"> 16</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00017"></a><span class="lineno"> 17</span> <span class="comment">@addtogroup grp_lda</span></div> |
| <div class="line"><a name="l00018"></a><span class="lineno"> 18</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00019"></a><span class="lineno"> 19</span> <span class="comment">@about</span></div> |
| <div class="line"><a name="l00020"></a><span class="lineno"> 20</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00021"></a><span class="lineno"> 21</span> <span class="comment">Latent Dirichlet Allocation (LDA) is an interesting generative probabilistic</span></div> |
| <div class="line"><a name="l00022"></a><span class="lineno"> 22</span> <span class="comment">model for natural texts and has received a lot of attention in recent years. </span></div> |
| <div class="line"><a name="l00023"></a><span class="lineno"> 23</span> <span class="comment">The model is quite versatile, having found uses in problems like automated </span></div> |
| <div class="line"><a name="l00024"></a><span class="lineno"> 24</span> <span class="comment">topic discovery, collaborative filtering, and document classification.</span></div> |
| <div class="line"><a name="l00025"></a><span class="lineno"> 25</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00026"></a><span class="lineno"> 26</span> <span class="comment">The LDA model posits that each document is associated with a mixture of various</span></div> |
| <div class="line"><a name="l00027"></a><span class="lineno"> 27</span> <span class="comment">topics (e.g. a document is related to Topic 1 with probability 0.7, and Topic 2</span></div> |
| <div class="line"><a name="l00028"></a><span class="lineno"> 28</span> <span class="comment">with probability 0.3), and that each word in the document is attributable to</span></div> |
| <div class="line"><a name="l00029"></a><span class="lineno"> 29</span> <span class="comment">one of the document's topics. There is a (symmetric) Dirichlet prior with</span></div> |
| <div class="line"><a name="l00030"></a><span class="lineno"> 30</span> <span class="comment">parameter \f$ \alpha \f$ on each document's topic mixture. In addition, there</span></div> |
| <div class="line"><a name="l00031"></a><span class="lineno"> 31</span> <span class="comment">is another (symmetric) Dirichlet prior with parameter \f$ \beta \f$ on the</span></div> |
| <div class="line"><a name="l00032"></a><span class="lineno"> 32</span> <span class="comment">distribution of words for each topic.</span></div> |
| <div class="line"><a name="l00033"></a><span class="lineno"> 33</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00034"></a><span class="lineno"> 34</span> <span class="comment">The following generative process then defines a distribution over a corpus of</span></div> |
| <div class="line"><a name="l00035"></a><span class="lineno"> 35</span> <span class="comment">documents. </span></div> |
| <div class="line"><a name="l00036"></a><span class="lineno"> 36</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00037"></a><span class="lineno"> 37</span> <span class="comment">- Sample for each topic \f$ i \f$, a per-topic word</span></div> |
| <div class="line"><a name="l00038"></a><span class="lineno"> 38</span> <span class="comment">distribution \f$ \phi_i \f$ from the Dirichlet(\f$\beta\f$) prior. </span></div> |
| <div class="line"><a name="l00039"></a><span class="lineno"> 39</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00040"></a><span class="lineno"> 40</span> <span class="comment">- For each document:</span></div> |
| <div class="line"><a name="l00041"></a><span class="lineno"> 41</span> <span class="comment"> - Sample a document length N from a suitable distribution, say, Poisson.</span></div> |
| <div class="line"><a name="l00042"></a><span class="lineno"> 42</span> <span class="comment"> - Sample a topic mixture \f$ \theta \f$ for the document from the</span></div> |
| <div class="line"><a name="l00043"></a><span class="lineno"> 43</span> <span class="comment">Dirichlet(\f$\alpha\f$) distribution. </span></div> |
| <div class="line"><a name="l00044"></a><span class="lineno"> 44</span> <span class="comment"> - For each of the N words:</span></div> |
| <div class="line"><a name="l00045"></a><span class="lineno"> 45</span> <span class="comment"> - Sample a topic \f$ z_n \f$ from the multinomial topic distribution \f$</span></div> |
| <div class="line"><a name="l00046"></a><span class="lineno"> 46</span> <span class="comment"> \theta \f$. </span></div> |
| <div class="line"><a name="l00047"></a><span class="lineno"> 47</span> <span class="comment"> - Sample a word \f$ w_n \f$ from the multinomial word distribution \f$</span></div> |
| <div class="line"><a name="l00048"></a><span class="lineno"> 48</span> <span class="comment"> \phi_{z_n} \f$ associated with topic \f$ z_n \f$.</span></div> |
| <div class="line"><a name="l00049"></a><span class="lineno"> 49</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00050"></a><span class="lineno"> 50</span> <span class="comment">In practice, only the words in each document are observable. The topic mixture</span></div> |
| <div class="line"><a name="l00051"></a><span class="lineno"> 51</span> <span class="comment">of each document and the topic for each word in each document are latent</span></div> |
| <div class="line"><a name="l00052"></a><span class="lineno"> 52</span> <span class="comment">unobservable variables that need to be inferred from the observables, and this</span></div> |
| <div class="line"><a name="l00053"></a><span class="lineno"> 53</span> <span class="comment">is the problem people refer to when they talk about the inference problem for</span></div> |
| <div class="line"><a name="l00054"></a><span class="lineno"> 54</span> <span class="comment">LDA. Exact inference is intractable, but several approximate inference</span></div> |
| <div class="line"><a name="l00055"></a><span class="lineno"> 55</span> <span class="comment">algorithms for LDA have been developed. The simple and effective Gibbs sampling</span></div> |
| <div class="line"><a name="l00056"></a><span class="lineno"> 56</span> <span class="comment">algorithm described in Griffiths and Steyvers [2] appears to be the current</span></div> |
| <div class="line"><a name="l00057"></a><span class="lineno"> 57</span> <span class="comment">algorithm of choice.</span></div> |
| <div class="line"><a name="l00058"></a><span class="lineno"> 58</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00059"></a><span class="lineno"> 59</span> <span class="comment">This implementation provides a parallel and scalable in-database solution for</span></div> |
| <div class="line"><a name="l00060"></a><span class="lineno"> 60</span> <span class="comment">LDA based on Gibbs sampling. Different with the implementations based on MPI or</span></div> |
| <div class="line"><a name="l00061"></a><span class="lineno"> 61</span> <span class="comment">Hadoop Map/Reduce, this implementation builds upon the shared-nothing MPP</span></div> |
| <div class="line"><a name="l00062"></a><span class="lineno"> 62</span> <span class="comment">databases and enables high-performance in-database analytics.</span></div> |
| <div class="line"><a name="l00063"></a><span class="lineno"> 63</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00064"></a><span class="lineno"> 64</span> <span class="comment">@input</span></div> |
| <div class="line"><a name="l00065"></a><span class="lineno"> 65</span> <span class="comment">The \b corpus/dataset to be analyzed is expected to be of the following form:</span></div> |
| <div class="line"><a name="l00066"></a><span class="lineno"> 66</span> <span class="comment"><pre>{TABLE|VIEW} <em>data_table</em> (</span></div> |
| <div class="line"><a name="l00067"></a><span class="lineno"> 67</span> <span class="comment"> <em>docid</em> INTEGER,</span></div> |
| <div class="line"><a name="l00068"></a><span class="lineno"> 68</span> <span class="comment"> <em>wordid</em> INTEGER,</span></div> |
| <div class="line"><a name="l00069"></a><span class="lineno"> 69</span> <span class="comment"> <em>count</em> INTEGER</span></div> |
| <div class="line"><a name="l00070"></a><span class="lineno"> 70</span> <span class="comment">)</pre></span></div> |
| <div class="line"><a name="l00071"></a><span class="lineno"> 71</span> <span class="comment">where \c docid refers to the document ID, \c wordid is the word ID (the index</span></div> |
| <div class="line"><a name="l00072"></a><span class="lineno"> 72</span> <span class="comment">of a word in the vocabulary), and \c count is the number of occurence of the</span></div> |
| <div class="line"><a name="l00073"></a><span class="lineno"> 73</span> <span class="comment">word in the document. </span></div> |
| <div class="line"><a name="l00074"></a><span class="lineno"> 74</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00075"></a><span class="lineno"> 75</span> <span class="comment">The \b vocabulary/dictionary that indexes all the words found in the corpus is</span></div> |
| <div class="line"><a name="l00076"></a><span class="lineno"> 76</span> <span class="comment">of the following form:</span></div> |
| <div class="line"><a name="l00077"></a><span class="lineno"> 77</span> <span class="comment"><pre>{TABLE|VIEW} <em>vocab_table</em> (</span></div> |
| <div class="line"><a name="l00078"></a><span class="lineno"> 78</span> <span class="comment"> <em>wordid</em> INTEGER,</span></div> |
| <div class="line"><a name="l00079"></a><span class="lineno"> 79</span> <span class="comment"> <em>word</em> TEXT,</span></div> |
| <div class="line"><a name="l00080"></a><span class="lineno"> 80</span> <span class="comment">)</pre></span></div> |
| <div class="line"><a name="l00081"></a><span class="lineno"> 81</span> <span class="comment">where \c wordid refers the word ID (the index of a word in the vocabulary) and</span></div> |
| <div class="line"><a name="l00082"></a><span class="lineno"> 82</span> <span class="comment">\c word is the actual word.</span></div> |
| <div class="line"><a name="l00083"></a><span class="lineno"> 83</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00084"></a><span class="lineno"> 84</span> <span class="comment">@usage</span></div> |
| <div class="line"><a name="l00085"></a><span class="lineno"> 85</span> <span class="comment">- The training (i.e. topic inference) can be done with the following function:</span></div> |
| <div class="line"><a name="l00086"></a><span class="lineno"> 86</span> <span class="comment"> <pre></span></div> |
| <div class="line"><a name="l00087"></a><span class="lineno"> 87</span> <span class="comment"> SELECT \ref lda_train(</span></div> |
| <div class="line"><a name="l00088"></a><span class="lineno"> 88</span> <span class="comment"> <em>'data_table'</em>,</span></div> |
| <div class="line"><a name="l00089"></a><span class="lineno"> 89</span> <span class="comment"> <em>'model_table'</em>,</span></div> |
| <div class="line"><a name="l00090"></a><span class="lineno"> 90</span> <span class="comment"> <em>'output_data_table'</em>, </span></div> |
| <div class="line"><a name="l00091"></a><span class="lineno"> 91</span> <span class="comment"> <em>voc_size</em>, </span></div> |
| <div class="line"><a name="l00092"></a><span class="lineno"> 92</span> <span class="comment"> <em>topic_num</em>,</span></div> |
| <div class="line"><a name="l00093"></a><span class="lineno"> 93</span> <span class="comment"> <em>iter_num</em>, </span></div> |
| <div class="line"><a name="l00094"></a><span class="lineno"> 94</span> <span class="comment"> <em>alpha</em>, </span></div> |
| <div class="line"><a name="l00095"></a><span class="lineno"> 95</span> <span class="comment"> <em>beta</em>)</span></div> |
| <div class="line"><a name="l00096"></a><span class="lineno"> 96</span> <span class="comment"> </pre></span></div> |
| <div class="line"><a name="l00097"></a><span class="lineno"> 97</span> <span class="comment"> </span></div> |
| <div class="line"><a name="l00098"></a><span class="lineno"> 98</span> <span class="comment"> This function stores the resulting model in <tt><em>model_table</em></tt>.</span></div> |
| <div class="line"><a name="l00099"></a><span class="lineno"> 99</span> <span class="comment"> The table has only 1 row and is in the following form:</span></div> |
| <div class="line"><a name="l00100"></a><span class="lineno"> 100</span> <span class="comment"> <pre>{TABLE} <em>model_table</em> (</span></div> |
| <div class="line"><a name="l00101"></a><span class="lineno"> 101</span> <span class="comment"> <em>voc_size</em> INTEGER,</span></div> |
| <div class="line"><a name="l00102"></a><span class="lineno"> 102</span> <span class="comment"> <em>topic_num</em> INTEGER,</span></div> |
| <div class="line"><a name="l00103"></a><span class="lineno"> 103</span> <span class="comment"> <em>alpha</em> FLOAT,</span></div> |
| <div class="line"><a name="l00104"></a><span class="lineno"> 104</span> <span class="comment"> <em>beta</em> FLOAT,</span></div> |
| <div class="line"><a name="l00105"></a><span class="lineno"> 105</span> <span class="comment"> <em>model</em> INTEGER[][])</span></div> |
| <div class="line"><a name="l00106"></a><span class="lineno"> 106</span> <span class="comment"> </pre></span></div> |
| <div class="line"><a name="l00107"></a><span class="lineno"> 107</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00108"></a><span class="lineno"> 108</span> <span class="comment"> This function also stores the topic counts and the topic assignments in</span></div> |
| <div class="line"><a name="l00109"></a><span class="lineno"> 109</span> <span class="comment"> each document in <tt><em>output_data_table</em></tt>. The table is in the</span></div> |
| <div class="line"><a name="l00110"></a><span class="lineno"> 110</span> <span class="comment"> following form:</span></div> |
| <div class="line"><a name="l00111"></a><span class="lineno"> 111</span> <span class="comment"> <pre>{TABLE} <em>output_data_table</em> (</span></div> |
| <div class="line"><a name="l00112"></a><span class="lineno"> 112</span> <span class="comment"> <em>docid</em> INTEGER,</span></div> |
| <div class="line"><a name="l00113"></a><span class="lineno"> 113</span> <span class="comment"> <em>wordcount</em> INTEGER,</span></div> |
| <div class="line"><a name="l00114"></a><span class="lineno"> 114</span> <span class="comment"> <em>words</em> INTEGER[],</span></div> |
| <div class="line"><a name="l00115"></a><span class="lineno"> 115</span> <span class="comment"> <em>counts</em> INTEGER[],</span></div> |
| <div class="line"><a name="l00116"></a><span class="lineno"> 116</span> <span class="comment"> <em>topic_count</em> INTEGER[],</span></div> |
| <div class="line"><a name="l00117"></a><span class="lineno"> 117</span> <span class="comment"> <em>topic_assignment</em> INTEGER[])</span></div> |
| <div class="line"><a name="l00118"></a><span class="lineno"> 118</span> <span class="comment"> </pre></span></div> |
| <div class="line"><a name="l00119"></a><span class="lineno"> 119</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00120"></a><span class="lineno"> 120</span> <span class="comment">- The prediction (i.e. labelling of test documents using a learned LDA model)</span></div> |
| <div class="line"><a name="l00121"></a><span class="lineno"> 121</span> <span class="comment"> can be done with the following function: </span></div> |
| <div class="line"><a name="l00122"></a><span class="lineno"> 122</span> <span class="comment"> <pre></span></div> |
| <div class="line"><a name="l00123"></a><span class="lineno"> 123</span> <span class="comment"> SELECT \ref lda_predict(</span></div> |
| <div class="line"><a name="l00124"></a><span class="lineno"> 124</span> <span class="comment"> <em>'data_table'</em>,</span></div> |
| <div class="line"><a name="l00125"></a><span class="lineno"> 125</span> <span class="comment"> <em>'model_table'</em>,</span></div> |
| <div class="line"><a name="l00126"></a><span class="lineno"> 126</span> <span class="comment"> <em>'output_table'</em>);</span></div> |
| <div class="line"><a name="l00127"></a><span class="lineno"> 127</span> <span class="comment"> </pre></span></div> |
| <div class="line"><a name="l00128"></a><span class="lineno"> 128</span> <span class="comment"> </span></div> |
| <div class="line"><a name="l00129"></a><span class="lineno"> 129</span> <span class="comment"> This function stores the prediction results in</span></div> |
| <div class="line"><a name="l00130"></a><span class="lineno"> 130</span> <span class="comment"> <tt><em>output_table</em></tt>. Each row in the table stores the topic</span></div> |
| <div class="line"><a name="l00131"></a><span class="lineno"> 131</span> <span class="comment"> distribution and the topic assignments for a docuemnt in the dataset. And</span></div> |
| <div class="line"><a name="l00132"></a><span class="lineno"> 132</span> <span class="comment"> the table is in the following form: </span></div> |
| <div class="line"><a name="l00133"></a><span class="lineno"> 133</span> <span class="comment"> <pre>{TABLE} <em>output_table</em> (</span></div> |
| <div class="line"><a name="l00134"></a><span class="lineno"> 134</span> <span class="comment"> <em>docid</em> INTEGER,</span></div> |
| <div class="line"><a name="l00135"></a><span class="lineno"> 135</span> <span class="comment"> <em>wordcount</em> INTEGER,</span></div> |
| <div class="line"><a name="l00136"></a><span class="lineno"> 136</span> <span class="comment"> <em>words</em> INTEGER,</span></div> |
| <div class="line"><a name="l00137"></a><span class="lineno"> 137</span> <span class="comment"> <em>counts</em> INTEGER,</span></div> |
| <div class="line"><a name="l00138"></a><span class="lineno"> 138</span> <span class="comment"> <em>topic_count</em> INTEGER[],</span></div> |
| <div class="line"><a name="l00139"></a><span class="lineno"> 139</span> <span class="comment"> <em>topic_assignment</em> INTEGER[])</span></div> |
| <div class="line"><a name="l00140"></a><span class="lineno"> 140</span> <span class="comment"> </pre></span></div> |
| <div class="line"><a name="l00141"></a><span class="lineno"> 141</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00142"></a><span class="lineno"> 142</span> <span class="comment">- This module also provides a function for computing the perplexity:</span></div> |
| <div class="line"><a name="l00143"></a><span class="lineno"> 143</span> <span class="comment"> <pre></span></div> |
| <div class="line"><a name="l00144"></a><span class="lineno"> 144</span> <span class="comment"> SELECT \ref lda_get_perplexity(</span></div> |
| <div class="line"><a name="l00145"></a><span class="lineno"> 145</span> <span class="comment"> <em>'model_table'</em>,</span></div> |
| <div class="line"><a name="l00146"></a><span class="lineno"> 146</span> <span class="comment"> <em>'output_data_table'</em>);</span></div> |
| <div class="line"><a name="l00147"></a><span class="lineno"> 147</span> <span class="comment"> </pre></span></div> |
| <div class="line"><a name="l00148"></a><span class="lineno"> 148</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00149"></a><span class="lineno"> 149</span> <span class="comment">@implementation</span></div> |
| <div class="line"><a name="l00150"></a><span class="lineno"> 150</span> <span class="comment">The input format for this module is very common in many machine learning</span></div> |
| <div class="line"><a name="l00151"></a><span class="lineno"> 151</span> <span class="comment">packages written in various lanugages, which allows users to generate</span></div> |
| <div class="line"><a name="l00152"></a><span class="lineno"> 152</span> <span class="comment">datasets using any existing document preprocessing tools or import existing</span></div> |
| <div class="line"><a name="l00153"></a><span class="lineno"> 153</span> <span class="comment">dataset very conveniently. Internally, the input data will be validated and then</span></div> |
| <div class="line"><a name="l00154"></a><span class="lineno"> 154</span> <span class="comment">converted to the following format for efficiency: </span></div> |
| <div class="line"><a name="l00155"></a><span class="lineno"> 155</span> <span class="comment"> <pre>{TABLE} <em>__internal_data_table__</em> (</span></div> |
| <div class="line"><a name="l00156"></a><span class="lineno"> 156</span> <span class="comment"> <em>docid</em> INTEGER,</span></div> |
| <div class="line"><a name="l00157"></a><span class="lineno"> 157</span> <span class="comment"> <em>wordcount</em> INTEGER,</span></div> |
| <div class="line"><a name="l00158"></a><span class="lineno"> 158</span> <span class="comment"> <em>words</em> INTEGER[],</span></div> |
| <div class="line"><a name="l00159"></a><span class="lineno"> 159</span> <span class="comment"> <em>counts</em> INTEGER[])</span></div> |
| <div class="line"><a name="l00160"></a><span class="lineno"> 160</span> <span class="comment"> </pre></span></div> |
| <div class="line"><a name="l00161"></a><span class="lineno"> 161</span> <span class="comment">where \c docid is the document ID, \c wordcount is the count of words in the</span></div> |
| <div class="line"><a name="l00162"></a><span class="lineno"> 162</span> <span class="comment">document, \c words is the list of unique words in the document, and \c counts</span></div> |
| <div class="line"><a name="l00163"></a><span class="lineno"> 163</span> <span class="comment">is the list of number of occurence of each unique word in the document. The</span></div> |
| <div class="line"><a name="l00164"></a><span class="lineno"> 164</span> <span class="comment">convertion can be done with the help of aggregation functions very easily.</span></div> |
| <div class="line"><a name="l00165"></a><span class="lineno"> 165</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00166"></a><span class="lineno"> 166</span> <span class="comment">@examp</span></div> |
| <div class="line"><a name="l00167"></a><span class="lineno"> 167</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00168"></a><span class="lineno"> 168</span> <span class="comment">We now give a usage example.</span></div> |
| <div class="line"><a name="l00169"></a><span class="lineno"> 169</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00170"></a><span class="lineno"> 170</span> <span class="comment">- As a first step, we need to prepare a dataset and vocabulary in the appropriate structure.</span></div> |
| <div class="line"><a name="l00171"></a><span class="lineno"> 171</span> <span class="comment"> \code</span></div> |
| <div class="line"><a name="l00172"></a><span class="lineno"> 172</span> <span class="comment"> CREATE TABLE my_vocab(wordid INT4, word TEXT)</span></div> |
| <div class="line"><a name="l00173"></a><span class="lineno"> 173</span> <span class="comment"> m4_ifdef(`__GREENPLUM__',`DISTRIBUTED BY (wordid)');</span></div> |
| <div class="line"><a name="l00174"></a><span class="lineno"> 174</span> <span class="comment"> </span></div> |
| <div class="line"><a name="l00175"></a><span class="lineno"> 175</span> <span class="comment"> INSERT INTO my_vocab VALUES</span></div> |
| <div class="line"><a name="l00176"></a><span class="lineno"> 176</span> <span class="comment"> (0, 'code'), (1, 'data'), (2, 'graph'), (3, 'image'), (4, 'input'), (5,</span></div> |
| <div class="line"><a name="l00177"></a><span class="lineno"> 177</span> <span class="comment"> 'layer'), (6, 'learner'), (7, 'loss'), (8, 'model'), (9, 'network'), (10,</span></div> |
| <div class="line"><a name="l00178"></a><span class="lineno"> 178</span> <span class="comment"> 'neuron'), (11, 'object'), (12, 'output'), (13, 'rate'), (14, 'set'), (15,</span></div> |
| <div class="line"><a name="l00179"></a><span class="lineno"> 179</span> <span class="comment"> 'signal'), (16, 'sparse'), (17, 'spatial'), (18, 'system'), (19, 'training');</span></div> |
| <div class="line"><a name="l00180"></a><span class="lineno"> 180</span> <span class="comment"> </span></div> |
| <div class="line"><a name="l00181"></a><span class="lineno"> 181</span> <span class="comment"> CREATE TABLE my_training </span></div> |
| <div class="line"><a name="l00182"></a><span class="lineno"> 182</span> <span class="comment"> (</span></div> |
| <div class="line"><a name="l00183"></a><span class="lineno"> 183</span> <span class="comment"> docid INT4, </span></div> |
| <div class="line"><a name="l00184"></a><span class="lineno"> 184</span> <span class="comment"> wordid INT4, </span></div> |
| <div class="line"><a name="l00185"></a><span class="lineno"> 185</span> <span class="comment"> count INT4</span></div> |
| <div class="line"><a name="l00186"></a><span class="lineno"> 186</span> <span class="comment"> )</span></div> |
| <div class="line"><a name="l00187"></a><span class="lineno"> 187</span> <span class="comment"> m4_ifdef(`__GREENPLUM__',`DISTRIBUTED BY (docid)');</span></div> |
| <div class="line"><a name="l00188"></a><span class="lineno"> 188</span> <span class="comment"> </span></div> |
| <div class="line"><a name="l00189"></a><span class="lineno"> 189</span> <span class="comment"> INSERT INTO my_training VALUES</span></div> |
| <div class="line"><a name="l00190"></a><span class="lineno"> 190</span> <span class="comment"> (0, 0, 2),(0, 3, 2),(0, 5, 1),(0, 7, 1),(0, 8, 1),(0, 9, 1),(0, 11, 1),(0, 13,</span></div> |
| <div class="line"><a name="l00191"></a><span class="lineno"> 191</span> <span class="comment"> 1), (1, 0, 1),(1, 3, 1),(1, 4, 1),(1, 5, 1),(1, 6, 1),(1, 7, 1),(1, 10, 1),(1,</span></div> |
| <div class="line"><a name="l00192"></a><span class="lineno"> 192</span> <span class="comment"> 14, 1),(1, 17, 1),(1, 18, 1), (2, 4, 2),(2, 5, 1),(2, 6, 2),(2, 12, 1),(2, 13,</span></div> |
| <div class="line"><a name="l00193"></a><span class="lineno"> 193</span> <span class="comment"> 1),(2, 15, 1),(2, 18, 2), (3, 0, 1),(3, 1, 2),(3, 12, 3),(3, 16, 1),(3, 17,</span></div> |
| <div class="line"><a name="l00194"></a><span class="lineno"> 194</span> <span class="comment"> 2),(3, 19, 1), (4, 1, 1),(4, 2, 1),(4, 3, 1),(4, 5, 1),(4, 6, 1),(4, 10, 1),(4,</span></div> |
| <div class="line"><a name="l00195"></a><span class="lineno"> 195</span> <span class="comment"> 11, 1),(4, 14, 1),(4, 18, 1),(4, 19, 1), (5, 0, 1),(5, 2, 1),(5, 5, 1),(5, 7,</span></div> |
| <div class="line"><a name="l00196"></a><span class="lineno"> 196</span> <span class="comment"> 1),(5, 10, 1),(5, 12, 1),(5, 16, 1),(5, 18, 1),(5, 19, 2), (6, 1, 1),(6, 3,</span></div> |
| <div class="line"><a name="l00197"></a><span class="lineno"> 197</span> <span class="comment"> 1),(6, 12, 2),(6, 13, 1),(6, 14, 2),(6, 15, 1),(6, 16, 1),(6, 17, 1), (7, 0,</span></div> |
| <div class="line"><a name="l00198"></a><span class="lineno"> 198</span> <span class="comment"> 1),(7, 2, 1),(7, 4, 1),(7, 5, 1),(7, 7, 2),(7, 8, 1),(7, 11, 1),(7, 14, 1),(7,</span></div> |
| <div class="line"><a name="l00199"></a><span class="lineno"> 199</span> <span class="comment"> 16, 1), (8, 2, 1),(8, 4, 4),(8, 6, 2),(8, 11, 1),(8, 15, 1),(8, 18, 1),</span></div> |
| <div class="line"><a name="l00200"></a><span class="lineno"> 200</span> <span class="comment"> (9, 0, 1),(9, 1, 1),(9, 4, 1),(9, 9, 2),(9, 12, 2),(9, 15, 1),(9, 18, 1),(9,</span></div> |
| <div class="line"><a name="l00201"></a><span class="lineno"> 201</span> <span class="comment"> 19, 1);</span></div> |
| <div class="line"><a name="l00202"></a><span class="lineno"> 202</span> <span class="comment"> </span></div> |
| <div class="line"><a name="l00203"></a><span class="lineno"> 203</span> <span class="comment"> </span></div> |
| <div class="line"><a name="l00204"></a><span class="lineno"> 204</span> <span class="comment"> CREATE TABLE my_testing </span></div> |
| <div class="line"><a name="l00205"></a><span class="lineno"> 205</span> <span class="comment"> (</span></div> |
| <div class="line"><a name="l00206"></a><span class="lineno"> 206</span> <span class="comment"> docid INT4, </span></div> |
| <div class="line"><a name="l00207"></a><span class="lineno"> 207</span> <span class="comment"> wordid INT4, </span></div> |
| <div class="line"><a name="l00208"></a><span class="lineno"> 208</span> <span class="comment"> count INT4</span></div> |
| <div class="line"><a name="l00209"></a><span class="lineno"> 209</span> <span class="comment"> )</span></div> |
| <div class="line"><a name="l00210"></a><span class="lineno"> 210</span> <span class="comment"> m4_ifdef(`__GREENPLUM__',`DISTRIBUTED BY (docid)');</span></div> |
| <div class="line"><a name="l00211"></a><span class="lineno"> 211</span> <span class="comment"> </span></div> |
| <div class="line"><a name="l00212"></a><span class="lineno"> 212</span> <span class="comment"> INSERT INTO my_testing VALUES</span></div> |
| <div class="line"><a name="l00213"></a><span class="lineno"> 213</span> <span class="comment"> (0, 0, 2),(0, 8, 1),(0, 9, 1),(0, 10, 1),(0, 12, 1),(0, 15, 2),(0, 18, 1),(0,</span></div> |
| <div class="line"><a name="l00214"></a><span class="lineno"> 214</span> <span class="comment"> 19, 1), (1, 0, 1),(1, 2, 1),(1, 5, 1),(1, 7, 1),(1, 12, 2),(1, 13, 1),(1, 16,</span></div> |
| <div class="line"><a name="l00215"></a><span class="lineno"> 215</span> <span class="comment"> 1),(1, 17, 1),(1, 18, 1), (2, 0, 1),(2, 1, 1),(2, 2, 1),(2, 3, 1),(2, 4, 1),(2,</span></div> |
| <div class="line"><a name="l00216"></a><span class="lineno"> 216</span> <span class="comment"> 5, 1),(2, 6, 1),(2, 12, 1),(2, 14, 1),(2, 18, 1), (3, 2, 2),(3, 6, 2),(3, 7,</span></div> |
| <div class="line"><a name="l00217"></a><span class="lineno"> 217</span> <span class="comment"> 1),(3, 9, 1),(3, 11, 2),(3, 14, 1),(3, 15, 1), (4, 1, 1),(4, 2, 2),(4, 3,</span></div> |
| <div class="line"><a name="l00218"></a><span class="lineno"> 218</span> <span class="comment"> 1),(4, 5, 2),(4, 6, 1),(4, 11, 1),(4, 18, 2);</span></div> |
| <div class="line"><a name="l00219"></a><span class="lineno"> 219</span> <span class="comment"> \endcode</span></div> |
| <div class="line"><a name="l00220"></a><span class="lineno"> 220</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00221"></a><span class="lineno"> 221</span> <span class="comment">- To perform training, we call the lda_train() function with the</span></div> |
| <div class="line"><a name="l00222"></a><span class="lineno"> 222</span> <span class="comment">appropriate parameters. Here is an example.</span></div> |
| <div class="line"><a name="l00223"></a><span class="lineno"> 223</span> <span class="comment"> \code</span></div> |
| <div class="line"><a name="l00224"></a><span class="lineno"> 224</span> <span class="comment"> SELECT MADLib.lda_train(</span></div> |
| <div class="line"><a name="l00225"></a><span class="lineno"> 225</span> <span class="comment"> 'my_training', 'my_model', 'my_outdata', 20, 5, 10, 5, 0.01);</span></div> |
| <div class="line"><a name="l00226"></a><span class="lineno"> 226</span> <span class="comment"> \endcode</span></div> |
| <div class="line"><a name="l00227"></a><span class="lineno"> 227</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00228"></a><span class="lineno"> 228</span> <span class="comment"> After a successful run of the lda_train() function, two tables will be</span></div> |
| <div class="line"><a name="l00229"></a><span class="lineno"> 229</span> <span class="comment"> generated, one for storing the learned models, and another for storing the</span></div> |
| <div class="line"><a name="l00230"></a><span class="lineno"> 230</span> <span class="comment"> output data table.</span></div> |
| <div class="line"><a name="l00231"></a><span class="lineno"> 231</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00232"></a><span class="lineno"> 232</span> <span class="comment"> To get the detailed information about the learned model, we can run the</span></div> |
| <div class="line"><a name="l00233"></a><span class="lineno"> 233</span> <span class="comment"> following commands:</span></div> |
| <div class="line"><a name="l00234"></a><span class="lineno"> 234</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00235"></a><span class="lineno"> 235</span> <span class="comment"> - The topic description by top-k words</span></div> |
| <div class="line"><a name="l00236"></a><span class="lineno"> 236</span> <span class="comment"> \code</span></div> |
| <div class="line"><a name="l00237"></a><span class="lineno"> 237</span> <span class="comment"> SELECT * FROM MADLib.lda_get_topic_desc(</span></div> |
| <div class="line"><a name="l00238"></a><span class="lineno"> 238</span> <span class="comment"> 'my_model', 'my_vocab', 'my_topic_desc', 15);</span></div> |
| <div class="line"><a name="l00239"></a><span class="lineno"> 239</span> <span class="comment"> \endcode</span></div> |
| <div class="line"><a name="l00240"></a><span class="lineno"> 240</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00241"></a><span class="lineno"> 241</span> <span class="comment"> - The per-topic word counts</span></div> |
| <div class="line"><a name="l00242"></a><span class="lineno"> 242</span> <span class="comment"> \code</span></div> |
| <div class="line"><a name="l00243"></a><span class="lineno"> 243</span> <span class="comment"> SELECT MADLib.lda_get_topic_word_count(</span></div> |
| <div class="line"><a name="l00244"></a><span class="lineno"> 244</span> <span class="comment"> 'my_model', 'my_topic_word_count');</span></div> |
| <div class="line"><a name="l00245"></a><span class="lineno"> 245</span> <span class="comment"> \endcode</span></div> |
| <div class="line"><a name="l00246"></a><span class="lineno"> 246</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00247"></a><span class="lineno"> 247</span> <span class="comment"> - The per-word topic counts</span></div> |
| <div class="line"><a name="l00248"></a><span class="lineno"> 248</span> <span class="comment"> \code</span></div> |
| <div class="line"><a name="l00249"></a><span class="lineno"> 249</span> <span class="comment"> SELECT MADLib.lda_get_word_topic_count(</span></div> |
| <div class="line"><a name="l00250"></a><span class="lineno"> 250</span> <span class="comment"> 'my_model', 'my_word_topic_count');</span></div> |
| <div class="line"><a name="l00251"></a><span class="lineno"> 251</span> <span class="comment"> \endcode</span></div> |
| <div class="line"><a name="l00252"></a><span class="lineno"> 252</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00253"></a><span class="lineno"> 253</span> <span class="comment"> To get the topic counts and the topic assignments for each doucment, we</span></div> |
| <div class="line"><a name="l00254"></a><span class="lineno"> 254</span> <span class="comment"> can run the following commands:</span></div> |
| <div class="line"><a name="l00255"></a><span class="lineno"> 255</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00256"></a><span class="lineno"> 256</span> <span class="comment"> - The per-document topic counts:</span></div> |
| <div class="line"><a name="l00257"></a><span class="lineno"> 257</span> <span class="comment"> \code</span></div> |
| <div class="line"><a name="l00258"></a><span class="lineno"> 258</span> <span class="comment"> SELECT </span></div> |
| <div class="line"><a name="l00259"></a><span class="lineno"> 259</span> <span class="comment"> docid, topic_count </span></div> |
| <div class="line"><a name="l00260"></a><span class="lineno"> 260</span> <span class="comment"> FROM my_outdata;</span></div> |
| <div class="line"><a name="l00261"></a><span class="lineno"> 261</span> <span class="comment"> \endcode</span></div> |
| <div class="line"><a name="l00262"></a><span class="lineno"> 262</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00263"></a><span class="lineno"> 263</span> <span class="comment"> - The per-document topic assignments:</span></div> |
| <div class="line"><a name="l00264"></a><span class="lineno"> 264</span> <span class="comment"> \code</span></div> |
| <div class="line"><a name="l00265"></a><span class="lineno"> 265</span> <span class="comment"> SELECT </span></div> |
| <div class="line"><a name="l00266"></a><span class="lineno"> 266</span> <span class="comment"> docid, words, counts, topic_assignment </span></div> |
| <div class="line"><a name="l00267"></a><span class="lineno"> 267</span> <span class="comment"> FROM my_outdata;</span></div> |
| <div class="line"><a name="l00268"></a><span class="lineno"> 268</span> <span class="comment"> \endcode</span></div> |
| <div class="line"><a name="l00269"></a><span class="lineno"> 269</span> <span class="comment"> By scanning \c words, \c counts, and \c topic_assignment together, we can</span></div> |
| <div class="line"><a name="l00270"></a><span class="lineno"> 270</span> <span class="comment"> get the topic assignment for each word in a document.</span></div> |
| <div class="line"><a name="l00271"></a><span class="lineno"> 271</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00272"></a><span class="lineno"> 272</span> <span class="comment">- To use a learned LDA model for prediction (i.e. to label new documents), we can use the following command: </span></div> |
| <div class="line"><a name="l00273"></a><span class="lineno"> 273</span> <span class="comment"> \code</span></div> |
| <div class="line"><a name="l00274"></a><span class="lineno"> 274</span> <span class="comment"> SELECT MADLib.lda_predict(</span></div> |
| <div class="line"><a name="l00275"></a><span class="lineno"> 275</span> <span class="comment"> 'my_testing', 'my_model', 'my_pred');</span></div> |
| <div class="line"><a name="l00276"></a><span class="lineno"> 276</span> <span class="comment"> \endcode</span></div> |
| <div class="line"><a name="l00277"></a><span class="lineno"> 277</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00278"></a><span class="lineno"> 278</span> <span class="comment"> After a successful run of the lda_predict() function, the prediction</span></div> |
| <div class="line"><a name="l00279"></a><span class="lineno"> 279</span> <span class="comment"> results will be generated and stored in <em>my_pred</em>. This table has</span></div> |
| <div class="line"><a name="l00280"></a><span class="lineno"> 280</span> <span class="comment"> the same schema as the <em>my_outdata</em> generated by the lda_train()</span></div> |
| <div class="line"><a name="l00281"></a><span class="lineno"> 281</span> <span class="comment"> function.</span></div> |
| <div class="line"><a name="l00282"></a><span class="lineno"> 282</span> <span class="comment"> </span></div> |
| <div class="line"><a name="l00283"></a><span class="lineno"> 283</span> <span class="comment"> To get te the topic counts and the topic assignments for each doucment, we</span></div> |
| <div class="line"><a name="l00284"></a><span class="lineno"> 284</span> <span class="comment"> can run the following commands:</span></div> |
| <div class="line"><a name="l00285"></a><span class="lineno"> 285</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00286"></a><span class="lineno"> 286</span> <span class="comment"> - The per-document topic counts:</span></div> |
| <div class="line"><a name="l00287"></a><span class="lineno"> 287</span> <span class="comment"> \code</span></div> |
| <div class="line"><a name="l00288"></a><span class="lineno"> 288</span> <span class="comment"> SELECT </span></div> |
| <div class="line"><a name="l00289"></a><span class="lineno"> 289</span> <span class="comment"> docid, topic_count </span></div> |
| <div class="line"><a name="l00290"></a><span class="lineno"> 290</span> <span class="comment"> FROM my_pred;</span></div> |
| <div class="line"><a name="l00291"></a><span class="lineno"> 291</span> <span class="comment"> \endcode</span></div> |
| <div class="line"><a name="l00292"></a><span class="lineno"> 292</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00293"></a><span class="lineno"> 293</span> <span class="comment"> - The per-document topic assignments:</span></div> |
| <div class="line"><a name="l00294"></a><span class="lineno"> 294</span> <span class="comment"> \code</span></div> |
| <div class="line"><a name="l00295"></a><span class="lineno"> 295</span> <span class="comment"> SELECT </span></div> |
| <div class="line"><a name="l00296"></a><span class="lineno"> 296</span> <span class="comment"> docid, words, counts, topic_assignment </span></div> |
| <div class="line"><a name="l00297"></a><span class="lineno"> 297</span> <span class="comment"> FROM my_pred;</span></div> |
| <div class="line"><a name="l00298"></a><span class="lineno"> 298</span> <span class="comment"> \endcode</span></div> |
| <div class="line"><a name="l00299"></a><span class="lineno"> 299</span> <span class="comment"> By scanning \c words, \c counts, and \c topic_assignment together, we can</span></div> |
| <div class="line"><a name="l00300"></a><span class="lineno"> 300</span> <span class="comment"> get the topic assignment for each word in a document.</span></div> |
| <div class="line"><a name="l00301"></a><span class="lineno"> 301</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00302"></a><span class="lineno"> 302</span> <span class="comment">- To compute the perplexity, we can use the following command:</span></div> |
| <div class="line"><a name="l00303"></a><span class="lineno"> 303</span> <span class="comment"> \code</span></div> |
| <div class="line"><a name="l00304"></a><span class="lineno"> 304</span> <span class="comment"> SELECT MADLib.lda_get_perplexity(</span></div> |
| <div class="line"><a name="l00305"></a><span class="lineno"> 305</span> <span class="comment"> 'my_model', 'my_pred');</span></div> |
| <div class="line"><a name="l00306"></a><span class="lineno"> 306</span> <span class="comment"> \endcode</span></div> |
| <div class="line"><a name="l00307"></a><span class="lineno"> 307</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00308"></a><span class="lineno"> 308</span> <span class="comment">@literature</span></div> |
| <div class="line"><a name="l00309"></a><span class="lineno"> 309</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00310"></a><span class="lineno"> 310</span> <span class="comment">[1] D.M. Blei, A.Y. Ng, M.I. Jordan, <em>Latent Dirichlet Allocation</em>,</span></div> |
| <div class="line"><a name="l00311"></a><span class="lineno"> 311</span> <span class="comment"> Journal of Machine Learning Research, vol. 3, pp. 993-1022, 2003.</span></div> |
| <div class="line"><a name="l00312"></a><span class="lineno"> 312</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00313"></a><span class="lineno"> 313</span> <span class="comment">[2] T. Griffiths and M. Steyvers, <em>Finding scientific topics</em>, PNAS,</span></div> |
| <div class="line"><a name="l00314"></a><span class="lineno"> 314</span> <span class="comment"> vol. 101, pp. 5228-5235, 2004.</span></div> |
| <div class="line"><a name="l00315"></a><span class="lineno"> 315</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00316"></a><span class="lineno"> 316</span> <span class="comment">[3] Y. Wang, H. Bai, M. Stanton, W-Y. Chen, and E.Y. Chang, <em>lda: Parallel</span></div> |
| <div class="line"><a name="l00317"></a><span class="lineno"> 317</span> <span class="comment"> Dirichlet Allocation for Large-scale Applications</em>, AAIM, 2009.</span></div> |
| <div class="line"><a name="l00318"></a><span class="lineno"> 318</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00319"></a><span class="lineno"> 319</span> <span class="comment">[4] http://en.wikipedia.org/wiki/Latent_Dirichlet_allocation</span></div> |
| <div class="line"><a name="l00320"></a><span class="lineno"> 320</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00321"></a><span class="lineno"> 321</span> <span class="comment">[5] J. Chang, Collapsed Gibbs sampling methods for topic models, R manual,</span></div> |
| <div class="line"><a name="l00322"></a><span class="lineno"> 322</span> <span class="comment"> 2010.</span></div> |
| <div class="line"><a name="l00323"></a><span class="lineno"> 323</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00324"></a><span class="lineno"> 324</span> <span class="comment">@sa File lda.sql_in documenting the SQL functions.</span></div> |
| <div class="line"><a name="l00325"></a><span class="lineno"> 325</span> <span class="comment">*/</span></div> |
| <div class="line"><a name="l00326"></a><span class="lineno"> 326</span> </div> |
| <div class="line"><a name="l00327"></a><span class="lineno"> 327</span> -- UDT for summarizing a UDF call</div> |
| <div class="line"><a name="l00328"></a><span class="lineno"> 328</span> DROP TYPE IF EXISTS MADLIB_SCHEMA.lda_result;</div> |
| <div class="line"><a name="l00329"></a><span class="lineno"> 329</span> CREATE TYPE MADLIB_SCHEMA.lda_result AS</div> |
| <div class="line"><a name="l00330"></a><span class="lineno"> 330</span> (</div> |
| <div class="line"><a name="l00331"></a><span class="lineno"> 331</span>  output_table TEXT,</div> |
| <div class="line"><a name="l00332"></a><span class="lineno"> 332</span>  description TEXT</div> |
| <div class="line"><a name="l00333"></a><span class="lineno"> 333</span> );</div> |
| <div class="line"><a name="l00334"></a><span class="lineno"> 334</span> <span class="comment"></span></div> |
| <div class="line"><a name="l00335"></a><span class="lineno"> 335</span> <span class="comment">/**</span></div> |
| <div class="line"><a name="l00336"></a><span class="lineno"> 336</span> <span class="comment"> * @brief This UDF provides an entry for the lda training process.</span></div> |
| <div class="line"><a name="l00337"></a><span class="lineno"> 337</span> <span class="comment"> * @param data_table Table storing the training dataset, each row is in</span></div> |
| <div class="line"><a name="l00338"></a><span class="lineno"> 338</span> <span class="comment"> * the form of <docid, wordid, count> where docid,</span></div> |
| <div class="line"><a name="l00339"></a><span class="lineno"> 339</span> <span class="comment"> * wordid, and count are all non-negative integers.</span></div> |
| <div class="line"><a name="l00340"></a><span class="lineno"> 340</span> <span class="comment"> * @param voc_size Size of the vocabulary (Note that the wordid should</span></div> |
| <div class="line"><a name="l00341"></a><span class="lineno"> 341</span> <span class="comment"> * be continous integers starting from 0 to voc_size -</span></div> |
| <div class="line"><a name="l00342"></a><span class="lineno"> 342</span> <span class="comment"> * 1. A data validation rountine will be called to</span></div> |
| <div class="line"><a name="l00343"></a><span class="lineno"> 343</span> <span class="comment"> * validate the dataset.)</span></div> |
| <div class="line"><a name="l00344"></a><span class="lineno"> 344</span> <span class="comment"> * @param topic_num Number of topics (e.g. 100)</span></div> |
| <div class="line"><a name="l00345"></a><span class="lineno"> 345</span> <span class="comment"> * @param iter_num Number of iterations (e.g. 60)</span></div> |
| <div class="line"><a name="l00346"></a><span class="lineno"> 346</span> <span class="comment"> * @param alpha Dirichlet parameter for the per-doc topic multinomial</span></div> |
| <div class="line"><a name="l00347"></a><span class="lineno"> 347</span> <span class="comment"> * (e.g. 50/topic_num)</span></div> |
| <div class="line"><a name="l00348"></a><span class="lineno"> 348</span> <span class="comment"> * @param beta Dirichlet parameter for the per-topic word multinomial</span></div> |
| <div class="line"><a name="l00349"></a><span class="lineno"> 349</span> <span class="comment"> * (e.g. 0.01)</span></div> |
| <div class="line"><a name="l00350"></a><span class="lineno"> 350</span> <span class="comment"> * @param model_table Table storing the learned models (voc_size, topic_num, </span></div> |
| <div class="line"><a name="l00351"></a><span class="lineno"> 351</span> <span class="comment"> * alpha, beta, per-word topic counts, and</span></div> |
| <div class="line"><a name="l00352"></a><span class="lineno"> 352</span> <span class="comment"> * corpus-level topic counts)</span></div> |
| <div class="line"><a name="l00353"></a><span class="lineno"> 353</span> <span class="comment"> * @param output_data_table Table storing the output data table in the form of</span></div> |
| <div class="line"><a name="l00354"></a><span class="lineno"> 354</span> <span class="comment"> * <docid, wordcount, words, counts, topic_count,</span></div> |
| <div class="line"><a name="l00355"></a><span class="lineno"> 355</span> <span class="comment"> * topic_assignment></span></div> |
| <div class="line"><a name="l00356"></a><span class="lineno"> 356</span> <span class="comment"> **/</span></div> |
| <div class="line"><a name="l00357"></a><span class="lineno"> 357</span> CREATE OR REPLACE FUNCTION</div> |
| <div class="line"><a name="l00358"></a><span class="lineno"> 358</span> MADLIB_SCHEMA.lda_train</div> |
| <div class="line"><a name="l00359"></a><span class="lineno"> 359</span> (</div> |
| <div class="line"><a name="l00360"></a><span class="lineno"> 360</span>  data_table TEXT, </div> |
| <div class="line"><a name="l00361"></a><span class="lineno"> 361</span>  model_table TEXT,</div> |
| <div class="line"><a name="l00362"></a><span class="lineno"> 362</span>  output_data_table TEXT,</div> |
| <div class="line"><a name="l00363"></a><span class="lineno"> 363</span>  voc_size INT4, </div> |
| <div class="line"><a name="l00364"></a><span class="lineno"> 364</span>  topic_num INT4, </div> |
| <div class="line"><a name="l00365"></a><span class="lineno"> 365</span>  iter_num INT4, </div> |
| <div class="line"><a name="l00366"></a><span class="lineno"> 366</span>  alpha FLOAT8, </div> |
| <div class="line"><a name="l00367"></a><span class="lineno"> 367</span>  beta FLOAT8 </div> |
| <div class="line"><a name="l00368"></a><span class="lineno"> 368</span> )</div> |
| <div class="line"><a name="l00369"></a><span class="lineno"> 369</span> RETURNS SETOF MADLIB_SCHEMA.lda_result AS $$</div> |
| <div class="line"><a name="l00370"></a><span class="lineno"> 370</span>  PythonFunctionBodyOnly(`lda', `lda<span class="stringliteral">')</span></div> |
| <div class="line"><a name="l00371"></a><span class="lineno"> 371</span> <span class="stringliteral"> lda.lda_train(</span></div> |
| <div class="line"><a name="l00372"></a><span class="lineno"> 372</span> <span class="stringliteral"> schema_madlib, data_table, model_table, output_data_table, voc_size,</span></div> |
| <div class="line"><a name="l00373"></a><span class="lineno"> 373</span> <span class="stringliteral"> topic_num, iter_num, alpha, beta</span></div> |
| <div class="line"><a name="l00374"></a><span class="lineno"> 374</span> <span class="stringliteral"> )</span></div> |
| <div class="line"><a name="l00375"></a><span class="lineno"> 375</span> <span class="stringliteral"> return [[model_table, '</span>model table<span class="stringliteral">'], </span></div> |
| <div class="line"><a name="l00376"></a><span class="lineno"> 376</span> <span class="stringliteral"> [output_data_table, '</span>output data table<span class="stringliteral">']]</span></div> |
| <div class="line"><a name="l00377"></a><span class="lineno"> 377</span> <span class="stringliteral">$$ LANGUAGE PLPYTHONU STRICT;</span></div> |
| <div class="line"><a name="l00378"></a><span class="lineno"> 378</span> <span class="stringliteral"></span></div> |
| <div class="line"><a name="l00379"></a><span class="lineno"> 379</span> <span class="stringliteral"></span><span class="comment"></span></div> |
| <div class="line"><a name="l00380"></a><span class="lineno"> 380</span> <span class="comment">/**</span></div> |
| <div class="line"><a name="l00381"></a><span class="lineno"> 381</span> <span class="comment"> * @brief This UDF provides an entry for the lda predicton process.</span></div> |
| <div class="line"><a name="l00382"></a><span class="lineno"> 382</span> <span class="comment"> * @param data_table Table storing the testing dataset, each row is in the </span></div> |
| <div class="line"><a name="l00383"></a><span class="lineno"> 383</span> <span class="comment"> * form of <docid, wordid, count></span></div> |
| <div class="line"><a name="l00384"></a><span class="lineno"> 384</span> <span class="comment"> * where docid, wordid, and count are all non-negative </span></div> |
| <div class="line"><a name="l00385"></a><span class="lineno"> 385</span> <span class="comment"> * integers.</span></div> |
| <div class="line"><a name="l00386"></a><span class="lineno"> 386</span> <span class="comment"> * @param model_table Table storing the learned models</span></div> |
| <div class="line"><a name="l00387"></a><span class="lineno"> 387</span> <span class="comment"> * @param output_table Table storing per-document topic counts and topic </span></div> |
| <div class="line"><a name="l00388"></a><span class="lineno"> 388</span> <span class="comment"> * assignments</span></div> |
| <div class="line"><a name="l00389"></a><span class="lineno"> 389</span> <span class="comment"> * @note default iter_num = 20</span></div> |
| <div class="line"><a name="l00390"></a><span class="lineno"> 390</span> <span class="comment"> **/</span></div> |
| <div class="line"><a name="l00391"></a><span class="lineno"> 391</span> CREATE OR REPLACE FUNCTION</div> |
| <div class="line"><a name="l00392"></a><span class="lineno"> 392</span> MADLIB_SCHEMA.lda_predict</div> |
| <div class="line"><a name="l00393"></a><span class="lineno"> 393</span> (</div> |
| <div class="line"><a name="l00394"></a><span class="lineno"> 394</span>  data_table TEXT,</div> |
| <div class="line"><a name="l00395"></a><span class="lineno"> 395</span>  model_table TEXT,</div> |
| <div class="line"><a name="l00396"></a><span class="lineno"> 396</span>  output_table TEXT</div> |
| <div class="line"><a name="l00397"></a><span class="lineno"> 397</span> )</div> |
| <div class="line"><a name="l00398"></a><span class="lineno"> 398</span> RETURNS SETOF MADLIB_SCHEMA.lda_result AS $$</div> |
| <div class="line"><a name="l00399"></a><span class="lineno"> 399</span>  PythonFunctionBodyOnly(`lda', `lda<span class="stringliteral">')</span></div> |
| <div class="line"><a name="l00400"></a><span class="lineno"> 400</span> <span class="stringliteral"> lda.lda_predict(schema_madlib, data_table, model_table, output_table)</span></div> |
| <div class="line"><a name="l00401"></a><span class="lineno"> 401</span> <span class="stringliteral"> return [[</span></div> |
| <div class="line"><a name="l00402"></a><span class="lineno"> 402</span> <span class="stringliteral"> output_table, </span></div> |
| <div class="line"><a name="l00403"></a><span class="lineno"> 403</span> <span class="stringliteral"> '</span>per-doc topic distribution and per-word topic assignments<span class="stringliteral">']]</span></div> |
| <div class="line"><a name="l00404"></a><span class="lineno"> 404</span> <span class="stringliteral">$$ LANGUAGE PLPYTHONU STRICT;</span></div> |
| <div class="line"><a name="l00405"></a><span class="lineno"> 405</span> <span class="stringliteral"></span><span class="comment"></span></div> |
| <div class="line"><a name="l00406"></a><span class="lineno"> 406</span> <span class="comment">/**</span></div> |
| <div class="line"><a name="l00407"></a><span class="lineno"> 407</span> <span class="comment"> * @brief A overloaded version which allows users to specify iter_num.</span></div> |
| <div class="line"><a name="l00408"></a><span class="lineno"> 408</span> <span class="comment"> **/</span></div> |
| <div class="line"><a name="l00409"></a><span class="lineno"> 409</span> CREATE OR REPLACE FUNCTION</div> |
| <div class="line"><a name="l00410"></a><span class="lineno"> 410</span> MADLIB_SCHEMA.lda_predict</div> |
| <div class="line"><a name="l00411"></a><span class="lineno"> 411</span> (</div> |
| <div class="line"><a name="l00412"></a><span class="lineno"> 412</span>  data_table TEXT,</div> |
| <div class="line"><a name="l00413"></a><span class="lineno"> 413</span>  model_table TEXT,</div> |
| <div class="line"><a name="l00414"></a><span class="lineno"> 414</span>  output_table TEXT,</div> |
| <div class="line"><a name="l00415"></a><span class="lineno"> 415</span>  iter_num INT4</div> |
| <div class="line"><a name="l00416"></a><span class="lineno"> 416</span> )</div> |
| <div class="line"><a name="l00417"></a><span class="lineno"> 417</span> RETURNS SETOF MADLIB_SCHEMA.lda_result AS $$</div> |
| <div class="line"><a name="l00418"></a><span class="lineno"> 418</span>  PythonFunctionBodyOnly(`lda', `lda<span class="stringliteral">')</span></div> |
| <div class="line"><a name="l00419"></a><span class="lineno"> 419</span> <span class="stringliteral"> lda.lda_predict(</span></div> |
| <div class="line"><a name="l00420"></a><span class="lineno"> 420</span> <span class="stringliteral"> schema_madlib, data_table, model_table, output_table, iter_num)</span></div> |
| <div class="line"><a name="l00421"></a><span class="lineno"> 421</span> <span class="stringliteral"> return [[</span></div> |
| <div class="line"><a name="l00422"></a><span class="lineno"> 422</span> <span class="stringliteral"> output_table, </span></div> |
| <div class="line"><a name="l00423"></a><span class="lineno"> 423</span> <span class="stringliteral"> '</span>per-doc topic distribution and per-word topic assignments<span class="stringliteral">']]</span></div> |
| <div class="line"><a name="l00424"></a><span class="lineno"> 424</span> <span class="stringliteral">$$ LANGUAGE PLPYTHONU STRICT;</span><span class="comment"></span></div> |
| <div class="line"><a name="l00425"></a><span class="lineno"> 425</span> <span class="comment">/**</span></div> |
| <div class="line"><a name="l00426"></a><span class="lineno"> 426</span> <span class="comment"> * @brief This UDF computes the per-topic word counts.</span></div> |
| <div class="line"><a name="l00427"></a><span class="lineno"> 427</span> <span class="comment"> * @param model_table The model table generated by the training process</span></div> |
| <div class="line"><a name="l00428"></a><span class="lineno"> 428</span> <span class="comment"> * @param output_table The output table storing the per-topic word counts</span></div> |
| <div class="line"><a name="l00429"></a><span class="lineno"> 429</span> <span class="comment"> **/</span></div> |
| <div class="line"><a name="l00430"></a><span class="lineno"> 430</span> CREATE OR REPLACE FUNCTION</div> |
| <div class="line"><a name="l00431"></a><span class="lineno"> 431</span> MADLIB_SCHEMA.lda_get_topic_word_count</div> |
| <div class="line"><a name="l00432"></a><span class="lineno"> 432</span> (</div> |
| <div class="line"><a name="l00433"></a><span class="lineno"> 433</span>  model_table TEXT,</div> |
| <div class="line"><a name="l00434"></a><span class="lineno"> 434</span>  output_table TEXT</div> |
| <div class="line"><a name="l00435"></a><span class="lineno"> 435</span> )</div> |
| <div class="line"><a name="l00436"></a><span class="lineno"> 436</span> RETURNS SETOF MADLIB_SCHEMA.lda_result AS $$</div> |
| <div class="line"><a name="l00437"></a><span class="lineno"> 437</span>  PythonFunctionBodyOnly(`lda', `lda<span class="stringliteral">')</span></div> |
| <div class="line"><a name="l00438"></a><span class="lineno"> 438</span> <span class="stringliteral"> lda.get_topic_word_count(schema_madlib, model_table, output_table)</span></div> |
| <div class="line"><a name="l00439"></a><span class="lineno"> 439</span> <span class="stringliteral"> return [[output_table, '</span>per-topic word counts<span class="stringliteral">']]</span></div> |
| <div class="line"><a name="l00440"></a><span class="lineno"> 440</span> <span class="stringliteral">$$ LANGUAGE plpythonu STRICT;</span></div> |
| <div class="line"><a name="l00441"></a><span class="lineno"> 441</span> <span class="stringliteral"></span><span class="comment"></span></div> |
| <div class="line"><a name="l00442"></a><span class="lineno"> 442</span> <span class="comment">/**</span></div> |
| <div class="line"><a name="l00443"></a><span class="lineno"> 443</span> <span class="comment"> * @brief This UDF computes the per-word topic counts.</span></div> |
| <div class="line"><a name="l00444"></a><span class="lineno"> 444</span> <span class="comment"> * @param model_table The model table generated by the training process</span></div> |
| <div class="line"><a name="l00445"></a><span class="lineno"> 445</span> <span class="comment"> * @param dist_table The output table storing the per-word topic counts</span></div> |
| <div class="line"><a name="l00446"></a><span class="lineno"> 446</span> <span class="comment"> **/</span></div> |
| <div class="line"><a name="l00447"></a><span class="lineno"> 447</span> CREATE OR REPLACE FUNCTION</div> |
| <div class="line"><a name="l00448"></a><span class="lineno"> 448</span> MADLIB_SCHEMA.lda_get_word_topic_count</div> |
| <div class="line"><a name="l00449"></a><span class="lineno"> 449</span> (</div> |
| <div class="line"><a name="l00450"></a><span class="lineno"> 450</span>  model_table TEXT, </div> |
| <div class="line"><a name="l00451"></a><span class="lineno"> 451</span>  output_table TEXT</div> |
| <div class="line"><a name="l00452"></a><span class="lineno"> 452</span> )</div> |
| <div class="line"><a name="l00453"></a><span class="lineno"> 453</span> RETURNS SETOF MADLIB_SCHEMA.lda_result AS $$</div> |
| <div class="line"><a name="l00454"></a><span class="lineno"> 454</span>  PythonFunctionBodyOnly(`lda', `lda<span class="stringliteral">')</span></div> |
| <div class="line"><a name="l00455"></a><span class="lineno"> 455</span> <span class="stringliteral"> lda.get_word_topic_count(schema_madlib, model_table, output_table)</span></div> |
| <div class="line"><a name="l00456"></a><span class="lineno"> 456</span> <span class="stringliteral"> return [[output_table, '</span>per-word topic counts<span class="stringliteral">']]</span></div> |
| <div class="line"><a name="l00457"></a><span class="lineno"> 457</span> <span class="stringliteral">$$ LANGUAGE plpythonu STRICT;</span></div> |
| <div class="line"><a name="l00458"></a><span class="lineno"> 458</span> <span class="stringliteral"></span><span class="comment"></span></div> |
| <div class="line"><a name="l00459"></a><span class="lineno"> 459</span> <span class="comment">/**</span></div> |
| <div class="line"><a name="l00460"></a><span class="lineno"> 460</span> <span class="comment"> * @brief This UDF gets the description for each topic (top-k words) </span></div> |
| <div class="line"><a name="l00461"></a><span class="lineno"> 461</span> <span class="comment"> * @param model_table The model table generated by the training process</span></div> |
| <div class="line"><a name="l00462"></a><span class="lineno"> 462</span> <span class="comment"> * @param vocab_table The vocabulary table (<wordid, word>)</span></div> |
| <div class="line"><a name="l00463"></a><span class="lineno"> 463</span> <span class="comment"> * @param top_k The number of top words for each topic description </span></div> |
| <div class="line"><a name="l00464"></a><span class="lineno"> 464</span> <span class="comment"> * @param desc_table The output table for storing the per-topic description </span></div> |
| <div class="line"><a name="l00465"></a><span class="lineno"> 465</span> <span class="comment"> **/</span></div> |
| <div class="line"><a name="l00466"></a><span class="lineno"> 466</span> CREATE OR REPLACE FUNCTION</div> |
| <div class="line"><a name="l00467"></a><span class="lineno"> 467</span> MADLIB_SCHEMA.lda_get_topic_desc</div> |
| <div class="line"><a name="l00468"></a><span class="lineno"> 468</span> (</div> |
| <div class="line"><a name="l00469"></a><span class="lineno"> 469</span>  model_table TEXT,</div> |
| <div class="line"><a name="l00470"></a><span class="lineno"> 470</span>  vocab_table TEXT,</div> |
| <div class="line"><a name="l00471"></a><span class="lineno"> 471</span>  desc_table TEXT,</div> |
| <div class="line"><a name="l00472"></a><span class="lineno"> 472</span>  top_k INT4</div> |
| <div class="line"><a name="l00473"></a><span class="lineno"> 473</span> )</div> |
| <div class="line"><a name="l00474"></a><span class="lineno"> 474</span> RETURNS SETOF MADLIB_SCHEMA.lda_result AS $$</div> |
| <div class="line"><a name="l00475"></a><span class="lineno"> 475</span>  PythonFunctionBodyOnly(`lda', `lda<span class="stringliteral">')</span></div> |
| <div class="line"><a name="l00476"></a><span class="lineno"> 476</span> <span class="stringliteral"> lda.get_topic_desc(</span></div> |
| <div class="line"><a name="l00477"></a><span class="lineno"> 477</span> <span class="stringliteral"> schema_madlib, model_table, vocab_table, desc_table, top_k)</span></div> |
| <div class="line"><a name="l00478"></a><span class="lineno"> 478</span> <span class="stringliteral"> return [[</span></div> |
| <div class="line"><a name="l00479"></a><span class="lineno"> 479</span> <span class="stringliteral"> desc_table, </span></div> |
| <div class="line"><a name="l00480"></a><span class="lineno"> 480</span> <span class="stringliteral"> """topic description, use "ORDER BY topicid, prob DESC" to check the</span></div> |
| <div class="line"><a name="l00481"></a><span class="lineno"> 481</span> <span class="stringliteral"> results"""]]</span></div> |
| <div class="line"><a name="l00482"></a><span class="lineno"> 482</span> <span class="stringliteral">$$ LANGUAGE plpythonu STRICT;</span></div> |
| <div class="line"><a name="l00483"></a><span class="lineno"> 483</span> <span class="stringliteral"></span><span class="comment"></span></div> |
| <div class="line"><a name="l00484"></a><span class="lineno"> 484</span> <span class="comment">/**</span></div> |
| <div class="line"><a name="l00485"></a><span class="lineno"> 485</span> <span class="comment"> * @brief This UDF assigns topics to words in a document randomly.</span></div> |
| <div class="line"><a name="l00486"></a><span class="lineno"> 486</span> <span class="comment"> * @param word_count The number of words in the document</span></div> |
| <div class="line"><a name="l00487"></a><span class="lineno"> 487</span> <span class="comment"> * @param topic_num The number of topics (specified by the user)</span></div> |
| <div class="line"><a name="l00488"></a><span class="lineno"> 488</span> <span class="comment"> * @return The topic counts and topic assignments </span></div> |
| <div class="line"><a name="l00489"></a><span class="lineno"> 489</span> <span class="comment"> **/</span></div> |
| <div class="line"><a name="l00490"></a><span class="lineno"> 490</span> CREATE OR REPLACE FUNCTION</div> |
| <div class="line"><a name="l00491"></a><span class="lineno"> 491</span> MADLIB_SCHEMA.__lda_random_assign</div> |
| <div class="line"><a name="l00492"></a><span class="lineno"> 492</span> (</div> |
| <div class="line"><a name="l00493"></a><span class="lineno"> 493</span>  word_count INT4, </div> |
| <div class="line"><a name="l00494"></a><span class="lineno"> 494</span>  topic_num INT4</div> |
| <div class="line"><a name="l00495"></a><span class="lineno"> 495</span> )</div> |
| <div class="line"><a name="l00496"></a><span class="lineno"> 496</span> RETURNS INT4[]</div> |
| <div class="line"><a name="l00497"></a><span class="lineno"> 497</span> AS 'MODULE_PATHNAME<span class="stringliteral">', '</span>lda_random_assign<span class="stringliteral">'</span></div> |
| <div class="line"><a name="l00498"></a><span class="lineno"> 498</span> <span class="stringliteral">LANGUAGE C STRICT;</span></div> |
| <div class="line"><a name="l00499"></a><span class="lineno"> 499</span> <span class="stringliteral"></span><span class="comment"></span></div> |
| <div class="line"><a name="l00500"></a><span class="lineno"> 500</span> <span class="comment">/**</span></div> |
| <div class="line"><a name="l00501"></a><span class="lineno"> 501</span> <span class="comment"> * @brief This UDF learns the topics of words in a document and is the main</span></div> |
| <div class="line"><a name="l00502"></a><span class="lineno"> 502</span> <span class="comment"> * step of a Gibbs sampling iteration. The model parameter (including the</span></div> |
| <div class="line"><a name="l00503"></a><span class="lineno"> 503</span> <span class="comment"> * per-word topic counts and corpus-level topic counts) is passed to this</span></div> |
| <div class="line"><a name="l00504"></a><span class="lineno"> 504</span> <span class="comment"> * function in the first call and then transfered to the rest calls through</span></div> |
| <div class="line"><a name="l00505"></a><span class="lineno"> 505</span> <span class="comment"> * fcinfo->flinfo->fn_extra to allow the immediate update.</span></div> |
| <div class="line"><a name="l00506"></a><span class="lineno"> 506</span> <span class="comment"> * @param words The set of unique words in the document</span></div> |
| <div class="line"><a name="l00507"></a><span class="lineno"> 507</span> <span class="comment"> * @param counts The counts of each unique words in the document</span></div> |
| <div class="line"><a name="l00508"></a><span class="lineno"> 508</span> <span class="comment"> * (sum(counts) = word_count)</span></div> |
| <div class="line"><a name="l00509"></a><span class="lineno"> 509</span> <span class="comment"> * @param doc_topic The current per-doc topic counts and topic</span></div> |
| <div class="line"><a name="l00510"></a><span class="lineno"> 510</span> <span class="comment"> * assignments</span></div> |
| <div class="line"><a name="l00511"></a><span class="lineno"> 511</span> <span class="comment"> * @param model The current model (including the per-word topic counts</span></div> |
| <div class="line"><a name="l00512"></a><span class="lineno"> 512</span> <span class="comment"> * and the corpus-level topic counts)</span></div> |
| <div class="line"><a name="l00513"></a><span class="lineno"> 513</span> <span class="comment"> * @param alpha The Dirichlet parameter for per-document topic multinomial </span></div> |
| <div class="line"><a name="l00514"></a><span class="lineno"> 514</span> <span class="comment"> * @param beta The Dirichlet parameter for per-topic word multinomial</span></div> |
| <div class="line"><a name="l00515"></a><span class="lineno"> 515</span> <span class="comment"> * @param voc_size The size of vocabulary</span></div> |
| <div class="line"><a name="l00516"></a><span class="lineno"> 516</span> <span class="comment"> * @param topic_num The number of topics</span></div> |
| <div class="line"><a name="l00517"></a><span class="lineno"> 517</span> <span class="comment"> * @return The learned topic counts and topic assignments </span></div> |
| <div class="line"><a name="l00518"></a><span class="lineno"> 518</span> <span class="comment"> **/</span></div> |
| <div class="line"><a name="l00519"></a><span class="lineno"> 519</span> CREATE OR REPLACE FUNCTION</div> |
| <div class="line"><a name="l00520"></a><span class="lineno"> 520</span> MADLIB_SCHEMA.__lda_gibbs_sample</div> |
| <div class="line"><a name="l00521"></a><span class="lineno"> 521</span> (</div> |
| <div class="line"><a name="l00522"></a><span class="lineno"> 522</span>  words INT4[], </div> |
| <div class="line"><a name="l00523"></a><span class="lineno"> 523</span>  counts INT4[],</div> |
| <div class="line"><a name="l00524"></a><span class="lineno"> 524</span>  doc_topic INT4[],</div> |
| <div class="line"><a name="l00525"></a><span class="lineno"> 525</span>  model INT4[], </div> |
| <div class="line"><a name="l00526"></a><span class="lineno"> 526</span>  alpha FLOAT8, </div> |
| <div class="line"><a name="l00527"></a><span class="lineno"> 527</span>  beta FLOAT8, </div> |
| <div class="line"><a name="l00528"></a><span class="lineno"> 528</span>  voc_size INT4, </div> |
| <div class="line"><a name="l00529"></a><span class="lineno"> 529</span>  topic_num INT4, </div> |
| <div class="line"><a name="l00530"></a><span class="lineno"> 530</span>  iter_num INT4</div> |
| <div class="line"><a name="l00531"></a><span class="lineno"> 531</span> )</div> |
| <div class="line"><a name="l00532"></a><span class="lineno"> 532</span> RETURNS INT4[]</div> |
| <div class="line"><a name="l00533"></a><span class="lineno"> 533</span> AS 'MODULE_PATHNAME<span class="stringliteral">', '</span>lda_gibbs_sample<span class="stringliteral">'</span></div> |
| <div class="line"><a name="l00534"></a><span class="lineno"> 534</span> <span class="stringliteral">LANGUAGE C;</span></div> |
| <div class="line"><a name="l00535"></a><span class="lineno"> 535</span> <span class="stringliteral"></span><span class="comment"></span></div> |
| <div class="line"><a name="l00536"></a><span class="lineno"> 536</span> <span class="comment">/**</span></div> |
| <div class="line"><a name="l00537"></a><span class="lineno"> 537</span> <span class="comment"> * @brief This UDF is the sfunc for the aggregator computing the topic counts</span></div> |
| <div class="line"><a name="l00538"></a><span class="lineno"> 538</span> <span class="comment"> * for each word and the topic count in the whole corpus. It scans the topic</span></div> |
| <div class="line"><a name="l00539"></a><span class="lineno"> 539</span> <span class="comment"> * assignments in a document and updates the topic counts.</span></div> |
| <div class="line"><a name="l00540"></a><span class="lineno"> 540</span> <span class="comment"> * @param state The topic counts</span></div> |
| <div class="line"><a name="l00541"></a><span class="lineno"> 541</span> <span class="comment"> * @param words The unique words in the document</span></div> |
| <div class="line"><a name="l00542"></a><span class="lineno"> 542</span> <span class="comment"> * @param counts The counts of each unique words in the document</span></div> |
| <div class="line"><a name="l00543"></a><span class="lineno"> 543</span> <span class="comment"> * (sum(counts) = word_count)</span></div> |
| <div class="line"><a name="l00544"></a><span class="lineno"> 544</span> <span class="comment"> * @param topic_assignment The topic assignments in the document</span></div> |
| <div class="line"><a name="l00545"></a><span class="lineno"> 545</span> <span class="comment"> * @param topic_num The number of topics</span></div> |
| <div class="line"><a name="l00546"></a><span class="lineno"> 546</span> <span class="comment"> * @return The updated state</span></div> |
| <div class="line"><a name="l00547"></a><span class="lineno"> 547</span> <span class="comment"> **/</span></div> |
| <div class="line"><a name="l00548"></a><span class="lineno"> 548</span> CREATE OR REPLACE FUNCTION </div> |
| <div class="line"><a name="l00549"></a><span class="lineno"> 549</span> MADLIB_SCHEMA.__lda_count_topic_sfunc</div> |
| <div class="line"><a name="l00550"></a><span class="lineno"> 550</span> (</div> |
| <div class="line"><a name="l00551"></a><span class="lineno"> 551</span>  state INT4[],</div> |
| <div class="line"><a name="l00552"></a><span class="lineno"> 552</span>  words INT4[], </div> |
| <div class="line"><a name="l00553"></a><span class="lineno"> 553</span>  counts INT4[],</div> |
| <div class="line"><a name="l00554"></a><span class="lineno"> 554</span>  topic_assignment INT4[], </div> |
| <div class="line"><a name="l00555"></a><span class="lineno"> 555</span>  voc_size INT4, </div> |
| <div class="line"><a name="l00556"></a><span class="lineno"> 556</span>  topic_num INT4</div> |
| <div class="line"><a name="l00557"></a><span class="lineno"> 557</span> )</div> |
| <div class="line"><a name="l00558"></a><span class="lineno"> 558</span> RETURNS INT4[]</div> |
| <div class="line"><a name="l00559"></a><span class="lineno"> 559</span> AS 'MODULE_PATHNAME<span class="stringliteral">', '</span>lda_count_topic_sfunc<span class="stringliteral">'</span></div> |
| <div class="line"><a name="l00560"></a><span class="lineno"> 560</span> <span class="stringliteral">LANGUAGE C;</span></div> |
| <div class="line"><a name="l00561"></a><span class="lineno"> 561</span> <span class="stringliteral"></span><span class="comment"></span></div> |
| <div class="line"><a name="l00562"></a><span class="lineno"> 562</span> <span class="comment">/**</span></div> |
| <div class="line"><a name="l00563"></a><span class="lineno"> 563</span> <span class="comment"> * @brief This UDF is the prefunc for the aggregator computing the per-word</span></div> |
| <div class="line"><a name="l00564"></a><span class="lineno"> 564</span> <span class="comment"> * topic counts.</span></div> |
| <div class="line"><a name="l00565"></a><span class="lineno"> 565</span> <span class="comment"> * @param state1 The local word topic counts</span></div> |
| <div class="line"><a name="l00566"></a><span class="lineno"> 566</span> <span class="comment"> * @param state2 The local word topic counts</span></div> |
| <div class="line"><a name="l00567"></a><span class="lineno"> 567</span> <span class="comment"> * @return The element-wise sum of two local states</span></div> |
| <div class="line"><a name="l00568"></a><span class="lineno"> 568</span> <span class="comment"> **/</span></div> |
| <div class="line"><a name="l00569"></a><span class="lineno"> 569</span> CREATE OR REPLACE FUNCTION </div> |
| <div class="line"><a name="l00570"></a><span class="lineno"> 570</span> MADLIB_SCHEMA.__lda_count_topic_prefunc</div> |
| <div class="line"><a name="l00571"></a><span class="lineno"> 571</span> (</div> |
| <div class="line"><a name="l00572"></a><span class="lineno"> 572</span>  state1 INT4[],</div> |
| <div class="line"><a name="l00573"></a><span class="lineno"> 573</span>  state2 INT4[]</div> |
| <div class="line"><a name="l00574"></a><span class="lineno"> 574</span> )</div> |
| <div class="line"><a name="l00575"></a><span class="lineno"> 575</span> RETURNS INT4[]</div> |
| <div class="line"><a name="l00576"></a><span class="lineno"> 576</span> AS 'MODULE_PATHNAME<span class="stringliteral">', '</span>lda_count_topic_prefunc<span class="stringliteral">'</span></div> |
| <div class="line"><a name="l00577"></a><span class="lineno"> 577</span> <span class="stringliteral">LANGUAGE C IMMUTABLE STRICT;</span></div> |
| <div class="line"><a name="l00578"></a><span class="lineno"> 578</span> <span class="stringliteral"></span><span class="comment"></span></div> |
| <div class="line"><a name="l00579"></a><span class="lineno"> 579</span> <span class="comment">/**</span></div> |
| <div class="line"><a name="l00580"></a><span class="lineno"> 580</span> <span class="comment"> * @brief This uda computes the word topic counts by scanning and summing</span></div> |
| <div class="line"><a name="l00581"></a><span class="lineno"> 581</span> <span class="comment"> * up topic assignments in each document.</span></div> |
| <div class="line"><a name="l00582"></a><span class="lineno"> 582</span> <span class="comment"> * @param words The unique words in the document</span></div> |
| <div class="line"><a name="l00583"></a><span class="lineno"> 583</span> <span class="comment"> * @param counts The counts of each unique words in the document</span></div> |
| <div class="line"><a name="l00584"></a><span class="lineno"> 584</span> <span class="comment"> * @param topic_assignment The topic assignments in the document</span></div> |
| <div class="line"><a name="l00585"></a><span class="lineno"> 585</span> <span class="comment"> * @param voc_size The size of vocabulary</span></div> |
| <div class="line"><a name="l00586"></a><span class="lineno"> 586</span> <span class="comment"> * @param topic_num The number of topics</span></div> |
| <div class="line"><a name="l00587"></a><span class="lineno"> 587</span> <span class="comment"> * @return The word topic counts (a 1-d array embeding a 2-d array)</span></div> |
| <div class="line"><a name="l00588"></a><span class="lineno"> 588</span> <span class="comment"> **/</span></div> |
| <div class="line"><a name="l00589"></a><span class="lineno"> 589</span> DROP AGGREGATE IF EXISTS </div> |
| <div class="line"><a name="l00590"></a><span class="lineno"> 590</span> MADLIB_SCHEMA.__lda_count_topic_agg</div> |
| <div class="line"><a name="l00591"></a><span class="lineno"> 591</span> (</div> |
| <div class="line"><a name="l00592"></a><span class="lineno"> 592</span>  INT4[], </div> |
| <div class="line"><a name="l00593"></a><span class="lineno"> 593</span>  INT4[], </div> |
| <div class="line"><a name="l00594"></a><span class="lineno"> 594</span>  INT4[], </div> |
| <div class="line"><a name="l00595"></a><span class="lineno"> 595</span>  INT4, </div> |
| <div class="line"><a name="l00596"></a><span class="lineno"> 596</span>  INT4</div> |
| <div class="line"><a name="l00597"></a><span class="lineno"> 597</span> );</div> |
| <div class="line"><a name="l00598"></a><span class="lineno"> 598</span> CREATE AGGREGATE </div> |
| <div class="line"><a name="l00599"></a><span class="lineno"> 599</span> MADLIB_SCHEMA.__lda_count_topic_agg</div> |
| <div class="line"><a name="l00600"></a><span class="lineno"> 600</span> (</div> |
| <div class="line"><a name="l00601"></a><span class="lineno"> 601</span>  INT4[], </div> |
| <div class="line"><a name="l00602"></a><span class="lineno"> 602</span>  INT4[], </div> |
| <div class="line"><a name="l00603"></a><span class="lineno"> 603</span>  INT4[], </div> |
| <div class="line"><a name="l00604"></a><span class="lineno"> 604</span>  INT4, </div> |
| <div class="line"><a name="l00605"></a><span class="lineno"> 605</span>  INT4</div> |
| <div class="line"><a name="l00606"></a><span class="lineno"> 606</span> )</div> |
| <div class="line"><a name="l00607"></a><span class="lineno"> 607</span> (</div> |
| <div class="line"><a name="l00608"></a><span class="lineno"> 608</span>  stype = INT4[],</div> |
| <div class="line"><a name="l00609"></a><span class="lineno"> 609</span>  sfunc = MADLIB_SCHEMA.__lda_count_topic_sfunc</div> |
| <div class="line"><a name="l00610"></a><span class="lineno"> 610</span>  m4_ifdef(</div> |
| <div class="line"><a name="l00611"></a><span class="lineno"> 611</span>  `GREENPLUM', </div> |
| <div class="line"><a name="l00612"></a><span class="lineno"> 612</span>  `, prefunc = MADLIB_SCHEMA.__lda_count_topic_prefunc<span class="stringliteral">'</span></div> |
| <div class="line"><a name="l00613"></a><span class="lineno"> 613</span> <span class="stringliteral"> )</span></div> |
| <div class="line"><a name="l00614"></a><span class="lineno"> 614</span> <span class="stringliteral">);</span></div> |
| <div class="line"><a name="l00615"></a><span class="lineno"> 615</span> <span class="stringliteral"></span><span class="comment"></span></div> |
| <div class="line"><a name="l00616"></a><span class="lineno"> 616</span> <span class="comment">/**</span></div> |
| <div class="line"><a name="l00617"></a><span class="lineno"> 617</span> <span class="comment"> * @brief This UDF computes the perplexity given the output data table and the</span></div> |
| <div class="line"><a name="l00618"></a><span class="lineno"> 618</span> <span class="comment"> * model table.</span></div> |
| <div class="line"><a name="l00619"></a><span class="lineno"> 619</span> <span class="comment"> * @param model_table The model table generated by lda_train</span></div> |
| <div class="line"><a name="l00620"></a><span class="lineno"> 620</span> <span class="comment"> * @param output_table The output data table generated by lda_predict</span></div> |
| <div class="line"><a name="l00621"></a><span class="lineno"> 621</span> <span class="comment"> * @return The perplexity</span></div> |
| <div class="line"><a name="l00622"></a><span class="lineno"> 622</span> <span class="comment"> **/</span></div> |
| <div class="line"><a name="l00623"></a><span class="lineno"> 623</span> CREATE OR REPLACE FUNCTION</div> |
| <div class="line"><a name="l00624"></a><span class="lineno"> 624</span> MADLIB_SCHEMA.lda_get_perplexity</div> |
| <div class="line"><a name="l00625"></a><span class="lineno"> 625</span> (</div> |
| <div class="line"><a name="l00626"></a><span class="lineno"> 626</span>  model_table TEXT,</div> |
| <div class="line"><a name="l00627"></a><span class="lineno"> 627</span>  output_data_table TEXT</div> |
| <div class="line"><a name="l00628"></a><span class="lineno"> 628</span> )</div> |
| <div class="line"><a name="l00629"></a><span class="lineno"> 629</span> RETURNS FLOAT8 AS $$</div> |
| <div class="line"><a name="l00630"></a><span class="lineno"> 630</span>  PythonFunctionBodyOnly(`lda', `lda<span class="stringliteral">')</span></div> |
| <div class="line"><a name="l00631"></a><span class="lineno"> 631</span> <span class="stringliteral"> return lda.get_perplexity(</span></div> |
| <div class="line"><a name="l00632"></a><span class="lineno"> 632</span> <span class="stringliteral"> schema_madlib, model_table, output_data_table)</span></div> |
| <div class="line"><a name="l00633"></a><span class="lineno"> 633</span> <span class="stringliteral">$$ LANGUAGE plpythonu STRICT;</span></div> |
| <div class="line"><a name="l00634"></a><span class="lineno"> 634</span> <span class="stringliteral"></span><span class="comment"></span></div> |
| <div class="line"><a name="l00635"></a><span class="lineno"> 635</span> <span class="comment">/**</span></div> |
| <div class="line"><a name="l00636"></a><span class="lineno"> 636</span> <span class="comment"> * @brief This UDF is the sfunc for the aggregator computing the perpleixty.</span></div> |
| <div class="line"><a name="l00637"></a><span class="lineno"> 637</span> <span class="comment"> * @param state The cached model plus perplexity</span></div> |
| <div class="line"><a name="l00638"></a><span class="lineno"> 638</span> <span class="comment"> * @param words The unique words in the document</span></div> |
| <div class="line"><a name="l00639"></a><span class="lineno"> 639</span> <span class="comment"> * @param counts The counts of each unique words in the document</span></div> |
| <div class="line"><a name="l00640"></a><span class="lineno"> 640</span> <span class="comment"> * @param doc_topic The topic counts in the document</span></div> |
| <div class="line"><a name="l00641"></a><span class="lineno"> 641</span> <span class="comment"> * @param model The learned model</span></div> |
| <div class="line"><a name="l00642"></a><span class="lineno"> 642</span> <span class="comment"> * @param alpha The Dirichlet parameter for per-document topic multinomial </span></div> |
| <div class="line"><a name="l00643"></a><span class="lineno"> 643</span> <span class="comment"> * @param beta The Dirichlet parameter for per-topic word multinomial</span></div> |
| <div class="line"><a name="l00644"></a><span class="lineno"> 644</span> <span class="comment"> * @param voc_size The size of vocabulary</span></div> |
| <div class="line"><a name="l00645"></a><span class="lineno"> 645</span> <span class="comment"> * @param topic_num The number of topics</span></div> |
| <div class="line"><a name="l00646"></a><span class="lineno"> 646</span> <span class="comment"> * @return The updated state </span></div> |
| <div class="line"><a name="l00647"></a><span class="lineno"> 647</span> <span class="comment"> **/</span></div> |
| <div class="line"><a name="l00648"></a><span class="lineno"> 648</span> CREATE OR REPLACE FUNCTION </div> |
| <div class="line"><a name="l00649"></a><span class="lineno"> 649</span> MADLIB_SCHEMA.__lda_perplexity_sfunc</div> |
| <div class="line"><a name="l00650"></a><span class="lineno"> 650</span> (</div> |
| <div class="line"><a name="l00651"></a><span class="lineno"> 651</span>  state INT4[],</div> |
| <div class="line"><a name="l00652"></a><span class="lineno"> 652</span>  words INT4[], </div> |
| <div class="line"><a name="l00653"></a><span class="lineno"> 653</span>  counts INT4[],</div> |
| <div class="line"><a name="l00654"></a><span class="lineno"> 654</span>  doc_topic INT4[], </div> |
| <div class="line"><a name="l00655"></a><span class="lineno"> 655</span>  model INT4[][], </div> |
| <div class="line"><a name="l00656"></a><span class="lineno"> 656</span>  alpha FLOAT8,</div> |
| <div class="line"><a name="l00657"></a><span class="lineno"> 657</span>  beta FLOAT8,</div> |
| <div class="line"><a name="l00658"></a><span class="lineno"> 658</span>  voc_size INT4, </div> |
| <div class="line"><a name="l00659"></a><span class="lineno"> 659</span>  topic_num INT4</div> |
| <div class="line"><a name="l00660"></a><span class="lineno"> 660</span> )</div> |
| <div class="line"><a name="l00661"></a><span class="lineno"> 661</span> RETURNS INT4[]</div> |
| <div class="line"><a name="l00662"></a><span class="lineno"> 662</span> AS 'MODULE_PATHNAME<span class="stringliteral">', '</span>lda_perplexity_sfunc<span class="stringliteral">'</span></div> |
| <div class="line"><a name="l00663"></a><span class="lineno"> 663</span> <span class="stringliteral">LANGUAGE C IMMUTABLE;</span></div> |
| <div class="line"><a name="l00664"></a><span class="lineno"> 664</span> <span class="stringliteral"></span><span class="comment"></span></div> |
| <div class="line"><a name="l00665"></a><span class="lineno"> 665</span> <span class="comment">/**</span></div> |
| <div class="line"><a name="l00666"></a><span class="lineno"> 666</span> <span class="comment"> * @brief This UDF is the prefunc for the aggregator computing the perplexity.</span></div> |
| <div class="line"><a name="l00667"></a><span class="lineno"> 667</span> <span class="comment"> * @param state1 The local state </span></div> |
| <div class="line"><a name="l00668"></a><span class="lineno"> 668</span> <span class="comment"> * @param state2 The local state </span></div> |
| <div class="line"><a name="l00669"></a><span class="lineno"> 669</span> <span class="comment"> * @return The merged state </span></div> |
| <div class="line"><a name="l00670"></a><span class="lineno"> 670</span> <span class="comment"> **/</span></div> |
| <div class="line"><a name="l00671"></a><span class="lineno"> 671</span> CREATE OR REPLACE FUNCTION </div> |
| <div class="line"><a name="l00672"></a><span class="lineno"> 672</span> MADLIB_SCHEMA.__lda_perplexity_prefunc</div> |
| <div class="line"><a name="l00673"></a><span class="lineno"> 673</span> (</div> |
| <div class="line"><a name="l00674"></a><span class="lineno"> 674</span>  state1 INT4[],</div> |
| <div class="line"><a name="l00675"></a><span class="lineno"> 675</span>  state2 INT4[] </div> |
| <div class="line"><a name="l00676"></a><span class="lineno"> 676</span> )</div> |
| <div class="line"><a name="l00677"></a><span class="lineno"> 677</span> RETURNS INT4[] </div> |
| <div class="line"><a name="l00678"></a><span class="lineno"> 678</span> AS 'MODULE_PATHNAME<span class="stringliteral">', '</span>lda_perplexity_prefunc<span class="stringliteral">'</span></div> |
| <div class="line"><a name="l00679"></a><span class="lineno"> 679</span> <span class="stringliteral">LANGUAGE C IMMUTABLE STRICT;</span></div> |
| <div class="line"><a name="l00680"></a><span class="lineno"> 680</span> <span class="stringliteral"></span><span class="comment"></span></div> |
| <div class="line"><a name="l00681"></a><span class="lineno"> 681</span> <span class="comment">/**</span></div> |
| <div class="line"><a name="l00682"></a><span class="lineno"> 682</span> <span class="comment"> * @brief This UDF is the finalfunc for the aggregator computing the perplexity.</span></div> |
| <div class="line"><a name="l00683"></a><span class="lineno"> 683</span> <span class="comment"> * @param state The merged state </span></div> |
| <div class="line"><a name="l00684"></a><span class="lineno"> 684</span> <span class="comment"> * @return The perpleixty</span></div> |
| <div class="line"><a name="l00685"></a><span class="lineno"> 685</span> <span class="comment"> **/</span></div> |
| <div class="line"><a name="l00686"></a><span class="lineno"> 686</span> CREATE OR REPLACE FUNCTION </div> |
| <div class="line"><a name="l00687"></a><span class="lineno"> 687</span> MADLIB_SCHEMA.__lda_perplexity_ffunc</div> |
| <div class="line"><a name="l00688"></a><span class="lineno"> 688</span> (</div> |
| <div class="line"><a name="l00689"></a><span class="lineno"> 689</span>  state INT4[]</div> |
| <div class="line"><a name="l00690"></a><span class="lineno"> 690</span> )</div> |
| <div class="line"><a name="l00691"></a><span class="lineno"> 691</span> RETURNS FLOAT8 </div> |
| <div class="line"><a name="l00692"></a><span class="lineno"> 692</span> AS 'MODULE_PATHNAME<span class="stringliteral">', '</span>lda_perplexity_ffunc<span class="stringliteral">'</span></div> |
| <div class="line"><a name="l00693"></a><span class="lineno"> 693</span> <span class="stringliteral">LANGUAGE C IMMUTABLE STRICT;</span></div> |
| <div class="line"><a name="l00694"></a><span class="lineno"> 694</span> <span class="stringliteral"></span><span class="comment"></span></div> |
| <div class="line"><a name="l00695"></a><span class="lineno"> 695</span> <span class="comment">/**</span></div> |
| <div class="line"><a name="l00696"></a><span class="lineno"> 696</span> <span class="comment"> * @brief This aggregator computes the perpleixty.</span></div> |
| <div class="line"><a name="l00697"></a><span class="lineno"> 697</span> <span class="comment"> * @param words The unique words in the document</span></div> |
| <div class="line"><a name="l00698"></a><span class="lineno"> 698</span> <span class="comment"> * @param counts The counts of each unique words in the document</span></div> |
| <div class="line"><a name="l00699"></a><span class="lineno"> 699</span> <span class="comment"> * @param doc_topic The topic counts in the document</span></div> |
| <div class="line"><a name="l00700"></a><span class="lineno"> 700</span> <span class="comment"> * @param model The learned model</span></div> |
| <div class="line"><a name="l00701"></a><span class="lineno"> 701</span> <span class="comment"> * @param alpha The Dirichlet parameter for per-document topic multinomial </span></div> |
| <div class="line"><a name="l00702"></a><span class="lineno"> 702</span> <span class="comment"> * @param beta The Dirichlet parameter for per-topic word multinomial</span></div> |
| <div class="line"><a name="l00703"></a><span class="lineno"> 703</span> <span class="comment"> * @param voc_size The size of vocabulary</span></div> |
| <div class="line"><a name="l00704"></a><span class="lineno"> 704</span> <span class="comment"> * @param topic_num The number of topics</span></div> |
| <div class="line"><a name="l00705"></a><span class="lineno"> 705</span> <span class="comment"> * @return The updated perplexity</span></div> |
| <div class="line"><a name="l00706"></a><span class="lineno"> 706</span> <span class="comment"> **/</span></div> |
| <div class="line"><a name="l00707"></a><span class="lineno"> 707</span> DROP AGGREGATE IF EXISTS </div> |
| <div class="line"><a name="l00708"></a><span class="lineno"> 708</span> MADLIB_SCHEMA.__lda_perplexity_agg</div> |
| <div class="line"><a name="l00709"></a><span class="lineno"> 709</span> (</div> |
| <div class="line"><a name="l00710"></a><span class="lineno"> 710</span>  INT4[], </div> |
| <div class="line"><a name="l00711"></a><span class="lineno"> 711</span>  INT4[],</div> |
| <div class="line"><a name="l00712"></a><span class="lineno"> 712</span>  INT4[], </div> |
| <div class="line"><a name="l00713"></a><span class="lineno"> 713</span>  INT4[], </div> |
| <div class="line"><a name="l00714"></a><span class="lineno"> 714</span>  FLOAT8,</div> |
| <div class="line"><a name="l00715"></a><span class="lineno"> 715</span>  FLOAT8,</div> |
| <div class="line"><a name="l00716"></a><span class="lineno"> 716</span>  INT4, </div> |
| <div class="line"><a name="l00717"></a><span class="lineno"> 717</span>  INT4</div> |
| <div class="line"><a name="l00718"></a><span class="lineno"> 718</span> );</div> |
| <div class="line"><a name="l00719"></a><span class="lineno"> 719</span> CREATE AGGREGATE </div> |
| <div class="line"><a name="l00720"></a><span class="lineno"> 720</span> MADLIB_SCHEMA.__lda_perplexity_agg</div> |
| <div class="line"><a name="l00721"></a><span class="lineno"> 721</span> (</div> |
| <div class="line"><a name="l00722"></a><span class="lineno"> 722</span>  INT4[], </div> |
| <div class="line"><a name="l00723"></a><span class="lineno"> 723</span>  INT4[],</div> |
| <div class="line"><a name="l00724"></a><span class="lineno"> 724</span>  INT4[], </div> |
| <div class="line"><a name="l00725"></a><span class="lineno"> 725</span>  INT4[], </div> |
| <div class="line"><a name="l00726"></a><span class="lineno"> 726</span>  FLOAT8,</div> |
| <div class="line"><a name="l00727"></a><span class="lineno"> 727</span>  FLOAT8,</div> |
| <div class="line"><a name="l00728"></a><span class="lineno"> 728</span>  INT4, </div> |
| <div class="line"><a name="l00729"></a><span class="lineno"> 729</span>  INT4</div> |
| <div class="line"><a name="l00730"></a><span class="lineno"> 730</span> )</div> |
| <div class="line"><a name="l00731"></a><span class="lineno"> 731</span> (</div> |
| <div class="line"><a name="l00732"></a><span class="lineno"> 732</span>  stype = INT4[],</div> |
| <div class="line"><a name="l00733"></a><span class="lineno"> 733</span>  sfunc = MADLIB_SCHEMA.__lda_perplexity_sfunc,</div> |
| <div class="line"><a name="l00734"></a><span class="lineno"> 734</span>  finalfunc = MADLIB_SCHEMA.__lda_perplexity_ffunc</div> |
| <div class="line"><a name="l00735"></a><span class="lineno"> 735</span>  m4_ifdef(</div> |
| <div class="line"><a name="l00736"></a><span class="lineno"> 736</span>  `GREENPLUM', </div> |
| <div class="line"><a name="l00737"></a><span class="lineno"> 737</span>  `, prefunc = MADLIB_SCHEMA.__lda_perplexity_prefunc<span class="stringliteral">'</span></div> |
| <div class="line"><a name="l00738"></a><span class="lineno"> 738</span> <span class="stringliteral"> )</span></div> |
| <div class="line"><a name="l00739"></a><span class="lineno"> 739</span> <span class="stringliteral">);</span></div> |
| <div class="line"><a name="l00740"></a><span class="lineno"> 740</span> <span class="stringliteral"></span><span class="comment"></span></div> |
| <div class="line"><a name="l00741"></a><span class="lineno"> 741</span> <span class="comment">/**</span></div> |
| <div class="line"><a name="l00742"></a><span class="lineno"> 742</span> <span class="comment"> * @brief Unnest a 2-D array into a set of 1-D arrays </span></div> |
| <div class="line"><a name="l00743"></a><span class="lineno"> 743</span> <span class="comment"> * @param arr The 2-D array to be unnested</span></div> |
| <div class="line"><a name="l00744"></a><span class="lineno"> 744</span> <span class="comment"> * @return The unnested 1-D arrays</span></div> |
| <div class="line"><a name="l00745"></a><span class="lineno"> 745</span> <span class="comment"> **/</span></div> |
| <div class="line"><a name="l00746"></a><span class="lineno"> 746</span> CREATE OR REPLACE FUNCTION </div> |
| <div class="line"><a name="l00747"></a><span class="lineno"> 747</span> MADLIB_SCHEMA.__lda_util_unnest</div> |
| <div class="line"><a name="l00748"></a><span class="lineno"> 748</span> (</div> |
| <div class="line"><a name="l00749"></a><span class="lineno"> 749</span>  arr INT4[][]</div> |
| <div class="line"><a name="l00750"></a><span class="lineno"> 750</span> )</div> |
| <div class="line"><a name="l00751"></a><span class="lineno"> 751</span> RETURNS SETOF INT4[] </div> |
| <div class="line"><a name="l00752"></a><span class="lineno"> 752</span> AS 'MODULE_PATHNAME<span class="stringliteral">', '</span>lda_unnest<span class="stringliteral">'</span></div> |
| <div class="line"><a name="l00753"></a><span class="lineno"> 753</span> <span class="stringliteral">LANGUAGE C IMMUTABLE STRICT;</span></div> |
| <div class="line"><a name="l00754"></a><span class="lineno"> 754</span> <span class="stringliteral"></span><span class="comment"></span></div> |
| <div class="line"><a name="l00755"></a><span class="lineno"> 755</span> <span class="comment">/**</span></div> |
| <div class="line"><a name="l00756"></a><span class="lineno"> 756</span> <span class="comment"> * @brief Transpose a 2-D array</span></div> |
| <div class="line"><a name="l00757"></a><span class="lineno"> 757</span> <span class="comment"> * @param matrix The input 2-D array</span></div> |
| <div class="line"><a name="l00758"></a><span class="lineno"> 758</span> <span class="comment"> * @param The transposed array</span></div> |
| <div class="line"><a name="l00759"></a><span class="lineno"> 759</span> <span class="comment"> **/</span></div> |
| <div class="line"><a name="l00760"></a><span class="lineno"> 760</span> CREATE OR REPLACE FUNCTION</div> |
| <div class="line"><a name="l00761"></a><span class="lineno"> 761</span> MADLIB_SCHEMA.__lda_util_transpose</div> |
| <div class="line"><a name="l00762"></a><span class="lineno"> 762</span> (</div> |
| <div class="line"><a name="l00763"></a><span class="lineno"> 763</span>  matrix INT4[][]</div> |
| <div class="line"><a name="l00764"></a><span class="lineno"> 764</span> )</div> |
| <div class="line"><a name="l00765"></a><span class="lineno"> 765</span> RETURNS INT4[][]</div> |
| <div class="line"><a name="l00766"></a><span class="lineno"> 766</span> AS 'MODULE_PATHNAME<span class="stringliteral">', '</span>lda_transpose<span class="stringliteral">'</span></div> |
| <div class="line"><a name="l00767"></a><span class="lineno"> 767</span> <span class="stringliteral">LANGUAGE C IMMUTABLE STRICT;</span></div> |
| <div class="line"><a name="l00768"></a><span class="lineno"> 768</span> <span class="stringliteral"></span><span class="comment"></span></div> |
| <div class="line"><a name="l00769"></a><span class="lineno"> 769</span> <span class="comment">/**</span></div> |
| <div class="line"><a name="l00770"></a><span class="lineno"> 770</span> <span class="comment"> * @brief L1 normalization with smoothing</span></div> |
| <div class="line"><a name="l00771"></a><span class="lineno"> 771</span> <span class="comment"> * @param arr The array to be normalized</span></div> |
| <div class="line"><a name="l00772"></a><span class="lineno"> 772</span> <span class="comment"> * @param smooth The smoothing parameter</span></div> |
| <div class="line"><a name="l00773"></a><span class="lineno"> 773</span> <span class="comment"> * @return The normalized vector</span></div> |
| <div class="line"><a name="l00774"></a><span class="lineno"> 774</span> <span class="comment"> **/</span></div> |
| <div class="line"><a name="l00775"></a><span class="lineno"> 775</span> CREATE OR REPLACE FUNCTION </div> |
| <div class="line"><a name="l00776"></a><span class="lineno"> 776</span> MADLIB_SCHEMA.__lda_util_norm_with_smoothing</div> |
| <div class="line"><a name="l00777"></a><span class="lineno"> 777</span> (</div> |
| <div class="line"><a name="l00778"></a><span class="lineno"> 778</span>  arr FLOAT8[],</div> |
| <div class="line"><a name="l00779"></a><span class="lineno"> 779</span>  smooth FLOAT8</div> |
| <div class="line"><a name="l00780"></a><span class="lineno"> 780</span> )</div> |
| <div class="line"><a name="l00781"></a><span class="lineno"> 781</span> RETURNS FLOAT8[] AS $$</div> |
| <div class="line"><a name="l00782"></a><span class="lineno"> 782</span>  PythonFunctionBodyOnly(`lda', `lda<span class="stringliteral">')</span></div> |
| <div class="line"><a name="l00783"></a><span class="lineno"> 783</span> <span class="stringliteral"> return lda.l1_norm_with_smoothing(arr, smooth)</span></div> |
| <div class="line"><a name="l00784"></a><span class="lineno"> 784</span> <span class="stringliteral">$$ LANGUAGE PLPYTHONU STRICT;</span></div> |
| <div class="line"><a name="l00785"></a><span class="lineno"> 785</span> <span class="stringliteral"></span><span class="comment"></span></div> |
| <div class="line"><a name="l00786"></a><span class="lineno"> 786</span> <span class="comment">/**</span></div> |
| <div class="line"><a name="l00787"></a><span class="lineno"> 787</span> <span class="comment"> * @brief This UDF returns the index of elements in a sorted order</span></div> |
| <div class="line"><a name="l00788"></a><span class="lineno"> 788</span> <span class="comment"> * @param arr The array to be sorted</span></div> |
| <div class="line"><a name="l00789"></a><span class="lineno"> 789</span> <span class="comment"> * @return The index of elements</span></div> |
| <div class="line"><a name="l00790"></a><span class="lineno"> 790</span> <span class="comment"> **/</span></div> |
| <div class="line"><a name="l00791"></a><span class="lineno"> 791</span> CREATE OR REPLACE FUNCTION</div> |
| <div class="line"><a name="l00792"></a><span class="lineno"> 792</span> MADLIB_SCHEMA.__lda_util_index_sort</div> |
| <div class="line"><a name="l00793"></a><span class="lineno"> 793</span> (</div> |
| <div class="line"><a name="l00794"></a><span class="lineno"> 794</span>  arr FLOAT8[]</div> |
| <div class="line"><a name="l00795"></a><span class="lineno"> 795</span> )</div> |
| <div class="line"><a name="l00796"></a><span class="lineno"> 796</span> RETURNS INT4[] AS $$</div> |
| <div class="line"><a name="l00797"></a><span class="lineno"> 797</span>  PythonFunctionBodyOnly(`lda', `lda<span class="stringliteral">')</span></div> |
| <div class="line"><a name="l00798"></a><span class="lineno"> 798</span> <span class="stringliteral"> return lda.index_sort(arr)</span></div> |
| <div class="line"><a name="l00799"></a><span class="lineno"> 799</span> <span class="stringliteral">$$ LANGUAGE plpythonu STRICT;</span></div> |
| <div class="line"><a name="l00800"></a><span class="lineno"> 800</span> <span class="stringliteral"></span><span class="comment"></span></div> |
| <div class="line"><a name="l00801"></a><span class="lineno"> 801</span> <span class="comment">/**</span></div> |
| <div class="line"><a name="l00802"></a><span class="lineno"> 802</span> <span class="comment"> * @brief This UDF checks the vocabulary and converts non-continous wordids into</span></div> |
| <div class="line"><a name="l00803"></a><span class="lineno"> 803</span> <span class="comment"> * continuous integers ranging from 0 to voc_size - 1.</span></div> |
| <div class="line"><a name="l00804"></a><span class="lineno"> 804</span> <span class="comment"> * @param vocab_table The vocabulary table in the form of </span></div> |
| <div class="line"><a name="l00805"></a><span class="lineno"> 805</span> <span class="comment"> <wordid::int4, word::text></span></div> |
| <div class="line"><a name="l00806"></a><span class="lineno"> 806</span> <span class="comment"> * @param output_vocab_table The regularized vocabulary table </span></div> |
| <div class="line"><a name="l00807"></a><span class="lineno"> 807</span> <span class="comment"> **/</span></div> |
| <div class="line"><a name="l00808"></a><span class="lineno"> 808</span> CREATE OR REPLACE FUNCTION</div> |
| <div class="line"><a name="l00809"></a><span class="lineno"> 809</span> MADLIB_SCHEMA.__lda_util_norm_vocab</div> |
| <div class="line"><a name="l00810"></a><span class="lineno"> 810</span> (</div> |
| <div class="line"><a name="l00811"></a><span class="lineno"> 811</span>  vocab_table TEXT,</div> |
| <div class="line"><a name="l00812"></a><span class="lineno"> 812</span>  output_vocab_table TEXT</div> |
| <div class="line"><a name="l00813"></a><span class="lineno"> 813</span> )</div> |
| <div class="line"><a name="l00814"></a><span class="lineno"> 814</span> RETURNS SETOF MADLIB_SCHEMA.lda_result AS $$</div> |
| <div class="line"><a name="l00815"></a><span class="lineno"> 815</span>  PythonFunctionBodyOnly(`lda', `lda<span class="stringliteral">')</span></div> |
| <div class="line"><a name="l00816"></a><span class="lineno"> 816</span> <span class="stringliteral"> lda.norm_vocab(vocab_table, output_vocab_table)</span></div> |
| <div class="line"><a name="l00817"></a><span class="lineno"> 817</span> <span class="stringliteral"> return [[output_vocab_table,'</span>normalized vocbulary table<span class="stringliteral">']]</span></div> |
| <div class="line"><a name="l00818"></a><span class="lineno"> 818</span> <span class="stringliteral">$$ LANGUAGE plpythonu STRICT;</span></div> |
| <div class="line"><a name="l00819"></a><span class="lineno"> 819</span> <span class="stringliteral"></span><span class="comment"></span></div> |
| <div class="line"><a name="l00820"></a><span class="lineno"> 820</span> <span class="comment">/**</span></div> |
| <div class="line"><a name="l00821"></a><span class="lineno"> 821</span> <span class="comment"> * @brief This UDF converts the data table according to the normalized</span></div> |
| <div class="line"><a name="l00822"></a><span class="lineno"> 822</span> <span class="comment"> * vocabulary, and all rows with non-positive count values will be removed</span></div> |
| <div class="line"><a name="l00823"></a><span class="lineno"> 823</span> <span class="comment"> * @param data_table The data table to be normalized</span></div> |
| <div class="line"><a name="l00824"></a><span class="lineno"> 824</span> <span class="comment"> * @param vocab_table The normalized vocabulary table</span></div> |
| <div class="line"><a name="l00825"></a><span class="lineno"> 825</span> <span class="comment"> * @param output_data_table The normalized data table</span></div> |
| <div class="line"><a name="l00826"></a><span class="lineno"> 826</span> <span class="comment"> **/</span></div> |
| <div class="line"><a name="l00827"></a><span class="lineno"> 827</span> CREATE OR REPLACE FUNCTION</div> |
| <div class="line"><a name="l00828"></a><span class="lineno"> 828</span> MADLIB_SCHEMA.__lda_util_norm_dataset</div> |
| <div class="line"><a name="l00829"></a><span class="lineno"> 829</span> (</div> |
| <div class="line"><a name="l00830"></a><span class="lineno"> 830</span>  data_table TEXT,</div> |
| <div class="line"><a name="l00831"></a><span class="lineno"> 831</span>  norm_vocab_table TEXT,</div> |
| <div class="line"><a name="l00832"></a><span class="lineno"> 832</span>  output_data_table TEXT</div> |
| <div class="line"><a name="l00833"></a><span class="lineno"> 833</span> )</div> |
| <div class="line"><a name="l00834"></a><span class="lineno"> 834</span> RETURNS SETOF MADLIB_SCHEMA.lda_result AS $$</div> |
| <div class="line"><a name="l00835"></a><span class="lineno"> 835</span>  PythonFunctionBodyOnly(`lda', `lda<span class="stringliteral">')</span></div> |
| <div class="line"><a name="l00836"></a><span class="lineno"> 836</span> <span class="stringliteral"> lda.norm_dataset(data_table, norm_vocab_table, output_data_table)</span></div> |
| <div class="line"><a name="l00837"></a><span class="lineno"> 837</span> <span class="stringliteral"> return [[output_data_table,'</span>normalized data table<span class="stringliteral">']]</span></div> |
| <div class="line"><a name="l00838"></a><span class="lineno"> 838</span> <span class="stringliteral">$$ LANGUAGE plpythonu STRICT;</span></div> |
| <div class="line"><a name="l00839"></a><span class="lineno"> 839</span> <span class="stringliteral"></span><span class="comment"></span></div> |
| <div class="line"><a name="l00840"></a><span class="lineno"> 840</span> <span class="comment">/**</span></div> |
| <div class="line"><a name="l00841"></a><span class="lineno"> 841</span> <span class="comment"> * @brief This UDF extracts the list of wordids from the data table and joins</span></div> |
| <div class="line"><a name="l00842"></a><span class="lineno"> 842</span> <span class="comment"> * it with the vocabulary table to get the list of common wordids, next it will</span></div> |
| <div class="line"><a name="l00843"></a><span class="lineno"> 843</span> <span class="comment"> * normalize the vocabulary based on the common wordids and then normalize the</span></div> |
| <div class="line"><a name="l00844"></a><span class="lineno"> 844</span> <span class="comment"> * data table based on the normalized vocabulary.</span></div> |
| <div class="line"><a name="l00845"></a><span class="lineno"> 845</span> <span class="comment"> * @param data_table The data table to be normalized</span></div> |
| <div class="line"><a name="l00846"></a><span class="lineno"> 846</span> <span class="comment"> * @param vocab_table The vocabulary table to be normalized</span></div> |
| <div class="line"><a name="l00847"></a><span class="lineno"> 847</span> <span class="comment"> * @param output_data_table The normalized data table</span></div> |
| <div class="line"><a name="l00848"></a><span class="lineno"> 848</span> <span class="comment"> * @param output_vocab_table The normalized vocabulary table</span></div> |
| <div class="line"><a name="l00849"></a><span class="lineno"> 849</span> <span class="comment"> **/</span></div> |
| <div class="line"><a name="l00850"></a><span class="lineno"> 850</span> CREATE OR REPLACE FUNCTION</div> |
| <div class="line"><a name="l00851"></a><span class="lineno"> 851</span> MADLIB_SCHEMA.__lda_util_conorm_data</div> |
| <div class="line"><a name="l00852"></a><span class="lineno"> 852</span> (</div> |
| <div class="line"><a name="l00853"></a><span class="lineno"> 853</span>  data_table TEXT,</div> |
| <div class="line"><a name="l00854"></a><span class="lineno"> 854</span>  vocab_table TEXT,</div> |
| <div class="line"><a name="l00855"></a><span class="lineno"> 855</span>  output_data_table TEXT,</div> |
| <div class="line"><a name="l00856"></a><span class="lineno"> 856</span>  output_vocab_table TEXT</div> |
| <div class="line"><a name="l00857"></a><span class="lineno"> 857</span> )</div> |
| <div class="line"><a name="l00858"></a><span class="lineno"> 858</span> RETURNS SETOF MADLIB_SCHEMA.lda_result AS $$</div> |
| <div class="line"><a name="l00859"></a><span class="lineno"> 859</span>  PythonFunctionBodyOnly(`lda', `lda<span class="stringliteral">')</span></div> |
| <div class="line"><a name="l00860"></a><span class="lineno"> 860</span> <span class="stringliteral"> lda.conorm_data(</span></div> |
| <div class="line"><a name="l00861"></a><span class="lineno"> 861</span> <span class="stringliteral"> data_table, vocab_table, output_data_table, output_vocab_table)</span></div> |
| <div class="line"><a name="l00862"></a><span class="lineno"> 862</span> <span class="stringliteral"> return [[output_data_table,'</span>normalized data table<span class="stringliteral">'],</span></div> |
| <div class="line"><a name="l00863"></a><span class="lineno"> 863</span> <span class="stringliteral"> [output_vocab_table,'</span>normalized vocab table<span class="stringliteral">']]</span></div> |
| <div class="line"><a name="l00864"></a><span class="lineno"> 864</span> <span class="stringliteral">$$ LANGUAGE plpythonu STRICT;</span></div> |
| <div class="line"><a name="l00865"></a><span class="lineno"> 865</span> <span class="stringliteral"></span></div> |
| </div><!-- fragment --></div><!-- contents --> |
| </div><!-- doc-content --> |
| <!-- start footer part --> |
| <div id="nav-path" class="navpath"><!-- id is needed for treeview function! --> |
| <ul> |
| <li class="navelem"><a class="el" href="dir_68267d1309a1af8e8297ef4c3efbcdba.html">src</a></li><li class="navelem"><a class="el" href="dir_efbcf68973d247bbf15f9eecae7f24e3.html">ports</a></li><li class="navelem"><a class="el" href="dir_a4a48839224ef8488facbffa8a397967.html">postgres</a></li><li class="navelem"><a class="el" href="dir_dc596537ad427a4d866006d1a3e1fe29.html">modules</a></li><li class="navelem"><a class="el" href="dir_6ff79b0655deb26abf8f86290b84a97c.html">lda</a></li><li class="navelem"><a class="el" href="lda_8sql__in.html">lda.sql_in</a></li> |
| <li class="footer">Generated on Tue Sep 10 2013 15:48:04 for MADlib by |
| <a href="http://www.doxygen.org/index.html"> |
| <img class="footer" src="doxygen.png" alt="doxygen"/></a> 1.8.4 </li> |
| </ul> |
| </div> |
| </body> |
| </html> |
