| <!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"/> |
| <title>MADlib: assoc_rules.sql_in Source File</title> |
| |
| <link href="tabs.css" rel="stylesheet" type="text/css"/> |
| <link href="doxygen.css" rel="stylesheet" type="text/css" /> |
| <link href="navtree.css" rel="stylesheet" type="text/css"/> |
| <script type="text/javascript" src="jquery.js"></script> |
| <script type="text/javascript" src="resize.js"></script> |
| <script type="text/javascript" src="navtree.js"></script> |
| <script type="text/javascript"> |
| $(document).ready(initResizable); |
| </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 src="../mathjax/MathJax.js"> |
| MathJax.Hub.Config({ |
| extensions: ["tex2jax.js", "TeX/AMSmath.js", "TeX/AMSsymbols.js"], |
| jax: ["input/TeX","output/HTML-CSS"], |
| }); |
| </script> |
| </head> |
| <body> |
| <div id="top"><!-- do not remove this div! --> |
| |
| |
| <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">0.6</span> <span style="font-size:10pt; font-style:italic"><a href="../latest/./assoc__rules_8sql__in_source.html"> A newer version is available</a></span> |
| </div> |
| <div id="projectbrief">User Documentation</div> |
| </td> |
| |
| |
| |
| </tr> |
| </tbody> |
| </table> |
| </div> |
| |
| <!-- Generated by Doxygen 1.7.5.1 --> |
| <script type="text/javascript"> |
| var searchBox = new SearchBox("searchBox", "search",false,'Search'); |
| </script> |
| <script type="text/javascript" src="dynsections.js"></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 class="current"><a href="files.html"><span>Files</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 id="navrow2" class="tabs2"> |
| <ul class="tablist"> |
| <li><a href="files.html"><span>File List</span></a></li> |
| <li><a href="globals.html"><span>File Members</span></a></li> |
| </ul> |
| </div> |
| </div> |
| <div id="side-nav" class="ui-resizable side-nav-resizable"> |
| <div id="nav-tree"> |
| <div id="nav-tree-contents"> |
| </div> |
| </div> |
| <div id="splitbar" style="-moz-user-select:none;" |
| class="ui-resizable-handle"> |
| </div> |
| </div> |
| <script type="text/javascript"> |
| initNavTree('assoc__rules_8sql__in.html',''); |
| </script> |
| <div id="doc-content"> |
| <div class="header"> |
| <div class="headertitle"> |
| <div class="title">assoc_rules.sql_in</div> </div> |
| </div> |
| <div class="contents"> |
| <a href="assoc__rules_8sql__in.html">Go to the documentation of this file.</a><div class="fragment"><pre class="fragment"><a name="l00001"></a>00001 <span class="comment">/* ----------------------------------------------------------------------- */</span><span class="comment">/**</span> |
| <a name="l00002"></a>00002 <span class="comment"> *</span> |
| <a name="l00003"></a>00003 <span class="comment"> * @file assoc_rules.sql_in</span> |
| <a name="l00004"></a>00004 <span class="comment"> *</span> |
| <a name="l00005"></a>00005 <span class="comment"> * @brief The \ref assoc_rules function computes association rules for a given set of data. The data is assumed to have two dimensions; items (between which we are trying to discover associations), and a transaction id. This tranaction id groups the items by event and could also be a user id, date, etc. depending on the context of the data. This function assumes the data is stored in two columns with one transaction id and one item per row.</span> |
| <a name="l00006"></a>00006 <span class="comment"></span> |
| <a name="l00007"></a>00007 <span class="comment"> * @date June 2011</span> |
| <a name="l00008"></a>00008 <span class="comment"> * @modified August 2012</span> |
| <a name="l00009"></a>00009 <span class="comment"> *</span> |
| <a name="l00010"></a>00010 <span class="comment"> * @sa For a brief introduction to the association rules implementation, see the module</span> |
| <a name="l00011"></a>00011 <span class="comment"> * description \ref grp_assoc_rules.</span> |
| <a name="l00012"></a>00012 <span class="comment"> *</span> |
| <a name="l00013"></a>00013 <span class="comment"> */</span><span class="comment">/* ----------------------------------------------------------------------- */</span> |
| <a name="l00014"></a>00014 |
| <a name="l00015"></a>00015 m4_include(`SQLCommon.m4<span class="stringliteral">')</span> |
| <a name="l00016"></a>00016 <span class="stringliteral"></span> |
| <a name="l00017"></a>00017 <span class="stringliteral"></span><span class="comment"></span> |
| <a name="l00018"></a>00018 <span class="comment">/**</span> |
| <a name="l00019"></a>00019 <span class="comment">@addtogroup grp_assoc_rules</span> |
| <a name="l00020"></a>00020 <span class="comment"></span> |
| <a name="l00021"></a>00021 <span class="comment">@about</span> |
| <a name="l00022"></a>00022 <span class="comment">This module implements the association rules data mining technique on a transactional data set. Given the names of a table and the columns, minimum support and confidence values, this function generates all single and multidimensional association rules that meet the minimum thresholds.</span> |
| <a name="l00023"></a>00023 <span class="comment"></span> |
| <a name="l00024"></a>00024 <span class="comment">Association rule mining is a widely used technique for discovering relationships between variables in a large data set (e.g items in a store that are commonly purchased together). The classic market basket analysis example using association rules is the "beer and diapers" rule. According to data mining urban legend, a study of customers' purchase behavior in a supermarket found that men often purchased beer and diapers together. After making this discovery, the managers strategically placed beer and diapers closer together on the shelves and saw a dramatic increase in sales. In addition to market basket analysis, association rules are also used in bioinformatics, web analytics, and several other fields.</span> |
| <a name="l00025"></a>00025 <span class="comment"></span> |
| <a name="l00026"></a>00026 <span class="comment">This type of data mining algorithm uses transactional data. Every transaction event has a unique identification, and each transaction consists of a set of items (or itemset). Purchases are considered binary (either it was purchased or not), and this implementation does not take into consideration the quantity of each item. For the MADlib association rules function, it is assumed that the data is stored in two columns with one item and transaction id per row. Transactions with multiple items will span multiple rows with one row per item.</span> |
| <a name="l00027"></a>00027 <span class="comment"></span> |
| <a name="l00028"></a>00028 <span class="comment"><pre></span> |
| <a name="l00029"></a>00029 <span class="comment"> tran_id | product</span> |
| <a name="l00030"></a>00030 <span class="comment"> ---------+---------</span> |
| <a name="l00031"></a>00031 <span class="comment"> 1 | 1</span> |
| <a name="l00032"></a>00032 <span class="comment"> 1 | 2</span> |
| <a name="l00033"></a>00033 <span class="comment"> 1 | 3</span> |
| <a name="l00034"></a>00034 <span class="comment"> 1 | 4</span> |
| <a name="l00035"></a>00035 <span class="comment"> 2 | 3</span> |
| <a name="l00036"></a>00036 <span class="comment"> 2 | 4</span> |
| <a name="l00037"></a>00037 <span class="comment"> 2 | 5</span> |
| <a name="l00038"></a>00038 <span class="comment"> 3 | 1</span> |
| <a name="l00039"></a>00039 <span class="comment"> 3 | 4</span> |
| <a name="l00040"></a>00040 <span class="comment"> 3 | 6</span> |
| <a name="l00041"></a>00041 <span class="comment"> ...</span> |
| <a name="l00042"></a>00042 <span class="comment"></pre></span> |
| <a name="l00043"></a>00043 <span class="comment"></span> |
| <a name="l00044"></a>00044 <span class="comment">\b Rules</span> |
| <a name="l00045"></a>00045 <span class="comment"></span> |
| <a name="l00046"></a>00046 <span class="comment">Association rules take the form "If X, then Y", where X and Y are non-empty itemsets. X and Y are called the antecedent and consequent, or the left-hand-side and right-hand-side, of the rule respectively. Using our previous example, the association rule may state "If {diapers}, then {beer}" with .2 support and .85 confidence.</span> |
| <a name="l00047"></a>00047 <span class="comment"></span> |
| <a name="l00048"></a>00048 <span class="comment">Given any association rule "If X, then Y", the association rules function will also calculate the following metrics:</span> |
| <a name="l00049"></a>00049 <span class="comment">- Support: The ratio of transactions that contain X to all transactions, T</span> |
| <a name="l00050"></a>00050 <span class="comment">\f[</span> |
| <a name="l00051"></a>00051 <span class="comment">S (X) = \frac{Total X}{Total transactions}</span> |
| <a name="l00052"></a>00052 <span class="comment">\f]</span> |
| <a name="l00053"></a>00053 <span class="comment"></span> |
| <a name="l00054"></a>00054 <span class="comment">- Confidence: The ratio of transactions that contain \f$ X,Y \f$ to transactions that contain \f$ X \f$. One could view this metric as the conditional probability of \f$ Y \f$ , given \f$ X \f$ . \f$ P(Y|X) \f$</span> |
| <a name="l00055"></a>00055 <span class="comment"></span> |
| <a name="l00056"></a>00056 <span class="comment">\f[</span> |
| <a name="l00057"></a>00057 <span class="comment">C (X \Rightarrow Y) = \frac{s(X \cap Y )}{s(X)}</span> |
| <a name="l00058"></a>00058 <span class="comment">\f]</span> |
| <a name="l00059"></a>00059 <span class="comment"></span> |
| <a name="l00060"></a>00060 <span class="comment">- Lift: The ratio of observed support of \f$ X,Y \f$ to the expected support of \f$ X,Y \f$ , assuming \f$ X \f$ and \f$ Y \f$ are independent.</span> |
| <a name="l00061"></a>00061 <span class="comment">\f[</span> |
| <a name="l00062"></a>00062 <span class="comment">L (X \Rightarrow Y) = \frac{s(X \cap Y )}{s(X) \cdot s(Y)}</span> |
| <a name="l00063"></a>00063 <span class="comment">\f]</span> |
| <a name="l00064"></a>00064 <span class="comment"></span> |
| <a name="l00065"></a>00065 <span class="comment">- Conviction: The ratio of expected support of \f$ X \f$ occurring without\f$ Y \f$ assuming \f$ X \f$ and \f$ \neg Y \f$ are independent, to the observed support of \f$ X \f$ occuring without \f$ Y \f$. If conviction is greater than 1, then this metric shows that incorrect predictions ( \f$ X \Rightarrow Y \f$ ) occur less often than if these two actions were independent. This metric can be viewed as the ratio that the association rule would be incorrect if the actions were independent (i.e. a conviction of 1.5 indicates that if the variables were independent, this rule would be incorrect 50% more often.)</span> |
| <a name="l00066"></a>00066 <span class="comment"></span> |
| <a name="l00067"></a>00067 <span class="comment">\f[</span> |
| <a name="l00068"></a>00068 <span class="comment">Conv (X \Rightarrow Y) = \frac{1 - S(Y)}{1 - C(X \Rightarrow Y)}</span> |
| <a name="l00069"></a>00069 <span class="comment">\f]</span> |
| <a name="l00070"></a>00070 <span class="comment"></span> |
| <a name="l00071"></a>00071 <span class="comment"></span> |
| <a name="l00072"></a>00072 <span class="comment">\b Apriori \b algorithm</span> |
| <a name="l00073"></a>00073 <span class="comment"></span> |
| <a name="l00074"></a>00074 <span class="comment">Although there are many algorithms that generate association rules, the classic algorithm used is called Apriori (which we implemented in this module). It is a breadth-first search, as opposed to depth-first searches like eclat. Frequent itemsets of order \f$ n \f$ are generated from sets of order \f$ n - 1 \f$. Using the downward closure property, all sets must have frequent subsets. There are two steps in this algorithm; generating frequent itemsets, and using these itemsets to construct the association rules. A simplified version of the algorithm is as follows, and assumes a minimum level of support and confidence is provided:</span> |
| <a name="l00075"></a>00075 <span class="comment"></span> |
| <a name="l00076"></a>00076 <span class="comment">\e Initial \e step</span> |
| <a name="l00077"></a>00077 <span class="comment">-# Generate all itemsets of order 1</span> |
| <a name="l00078"></a>00078 <span class="comment">-# Eliminate itemsets that have support is less than minimum support</span> |
| <a name="l00079"></a>00079 <span class="comment"></span> |
| <a name="l00080"></a>00080 <span class="comment">\e Main \e algorithm</span> |
| <a name="l00081"></a>00081 <span class="comment">-# For \f$ n \ge 2 \f$, generate itemsets of order \f$ n \f$ by combining the itemsets of order \f$ n - 1 \f$. This is done by doing the union of two itemsets that have identical items except one.</span> |
| <a name="l00082"></a>00082 <span class="comment">-# Eliminate itemsets that have (n-1) order subsets with insufficient support</span> |
| <a name="l00083"></a>00083 <span class="comment">-# Eliminate itemsets with insufficient support</span> |
| <a name="l00084"></a>00084 <span class="comment">-# Repeat until itemsets cannot be generated</span> |
| <a name="l00085"></a>00085 <span class="comment"></span> |
| <a name="l00086"></a>00086 <span class="comment">\e Association \e rule \e generation</span> |
| <a name="l00087"></a>00087 <span class="comment"></span> |
| <a name="l00088"></a>00088 <span class="comment">Given a frequent itemset \f$ A \f$ generated from the Apriori algorithm, and all subsets \f$ B \f$ , we generate rules such that \f$ B \Rightarrow (A - B) \f$ meets minimum confidence requirements.</span> |
| <a name="l00089"></a>00089 <span class="comment"></span> |
| <a name="l00090"></a>00090 <span class="comment">@input</span> |
| <a name="l00091"></a>00091 <span class="comment"></span> |
| <a name="l00092"></a>00092 <span class="comment">The input data is expected to be of the following form:</span> |
| <a name="l00093"></a>00093 <span class="comment"><pre>{TABLE|VIEW} <em>input_table</em> (</span> |
| <a name="l00094"></a>00094 <span class="comment"> <em>trans_id</em> INTEGER,</span> |
| <a name="l00095"></a>00095 <span class="comment"> <em>product</em> TEXT</span> |
| <a name="l00096"></a>00096 <span class="comment">)</pre></span> |
| <a name="l00097"></a>00097 <span class="comment"></span> |
| <a name="l00098"></a>00098 <span class="comment">The algorithm will map the product names to consective integer ids starting at 1. If they are already structured this way, then the ids will not change.</span> |
| <a name="l00099"></a>00099 <span class="comment"></span> |
| <a name="l00100"></a>00100 <span class="comment">@usage</span> |
| <a name="l00101"></a>00101 <span class="comment">- Association rules can be called by:</span> |
| <a name="l00102"></a>00102 <span class="comment"> <pre>SELECT \ref assoc_rules(</span> |
| <a name="l00103"></a>00103 <span class="comment"> <em>support</em>, <em>confidence</em>,'<em>tid_col</em>','<em>item_col</em>',</span> |
| <a name="l00104"></a>00104 <span class="comment"> '<em>input_table</em>','<em>output_schema</em>', <em> verbose </em></span> |
| <a name="l00105"></a>00105 <span class="comment"> );</pre></span> |
| <a name="l00106"></a>00106 <span class="comment"> This will generate all association rules that meet a minimum support of <em>support</em> and confidence of <em>confidence</em>.</span> |
| <a name="l00107"></a>00107 <span class="comment"></span> |
| <a name="l00108"></a>00108 <span class="comment">- The results containing the rules, support, confidence, lift, and conviction are stored in the table assoc_rules in the schema specified by <em>output_schema</em>.</span> |
| <a name="l00109"></a>00109 <span class="comment"><pre></span> |
| <a name="l00110"></a>00110 <span class="comment"> Table "output_schema.assoc_rules"</span> |
| <a name="l00111"></a>00111 <span class="comment"> Column | Type | Modifiers </span> |
| <a name="l00112"></a>00112 <span class="comment"> ------------+------------------+-----------</span> |
| <a name="l00113"></a>00113 <span class="comment"> ruleid | integer | </span> |
| <a name="l00114"></a>00114 <span class="comment"> pre | text[] | </span> |
| <a name="l00115"></a>00115 <span class="comment"> post | text[] | </span> |
| <a name="l00116"></a>00116 <span class="comment"> support | double precision | </span> |
| <a name="l00117"></a>00117 <span class="comment"> confidence | double precision | </span> |
| <a name="l00118"></a>00118 <span class="comment"> lift | double precision | </span> |
| <a name="l00119"></a>00119 <span class="comment"> conviction | double precision | </span> |
| <a name="l00120"></a>00120 <span class="comment"> Distributed by: (ruleid)</span> |
| <a name="l00121"></a>00121 <span class="comment"></span> |
| <a name="l00122"></a>00122 <span class="comment"></pre></span> |
| <a name="l00123"></a>00123 <span class="comment"> The \c pre and \c post are the itemsets of left and right hand sides of the association rule respectively. The \c support, \c confidence, \c lift, and \c conviction columns are calculated as mentioned in the about section.</span> |
| <a name="l00124"></a>00124 <span class="comment"></span> |
| <a name="l00125"></a>00125 <span class="comment">@implementation</span> |
| <a name="l00126"></a>00126 <span class="comment"></span> |
| <a name="l00127"></a>00127 <span class="comment">The association rules function will always create a table named assoc_rules. Please make a copy of this table before running the function again if you would like to keep multiple association rule tables.</span> |
| <a name="l00128"></a>00128 <span class="comment"></span> |
| <a name="l00129"></a>00129 <span class="comment">@examp</span> |
| <a name="l00130"></a>00130 <span class="comment"></span> |
| <a name="l00131"></a>00131 <span class="comment">Let us take a look at some sample transactional data and generate association rules:</span> |
| <a name="l00132"></a>00132 <span class="comment"></span> |
| <a name="l00133"></a>00133 <span class="comment">\code</span> |
| <a name="l00134"></a>00134 <span class="comment">DROP TABLE IF EXISTS test_data;</span> |
| <a name="l00135"></a>00135 <span class="comment">CREATE TABLE test_data (</span> |
| <a name="l00136"></a>00136 <span class="comment"> trans_id INT</span> |
| <a name="l00137"></a>00137 <span class="comment"> , product text</span> |
| <a name="l00138"></a>00138 <span class="comment">);</span> |
| <a name="l00139"></a>00139 <span class="comment"></span> |
| <a name="l00140"></a>00140 <span class="comment">INSERT INTO test_data VALUES (1, 'beer');</span> |
| <a name="l00141"></a>00141 <span class="comment">INSERT INTO test_data VALUES (1, 'diapers');</span> |
| <a name="l00142"></a>00142 <span class="comment">INSERT INTO test_data VALUES (1, 'chips');</span> |
| <a name="l00143"></a>00143 <span class="comment">INSERT INTO test_data VALUES (2, 'beer');</span> |
| <a name="l00144"></a>00144 <span class="comment">INSERT INTO test_data VALUES (2, 'diapers');</span> |
| <a name="l00145"></a>00145 <span class="comment">INSERT INTO test_data VALUES (3, 'beer');</span> |
| <a name="l00146"></a>00146 <span class="comment">INSERT INTO test_data VALUES (3, 'diapers');</span> |
| <a name="l00147"></a>00147 <span class="comment">INSERT INTO test_data VALUES (4, 'beer');</span> |
| <a name="l00148"></a>00148 <span class="comment">INSERT INTO test_data VALUES (4, 'chips');</span> |
| <a name="l00149"></a>00149 <span class="comment">INSERT INTO test_data VALUES (5, 'beer');</span> |
| <a name="l00150"></a>00150 <span class="comment">INSERT INTO test_data VALUES (6, 'beer');</span> |
| <a name="l00151"></a>00151 <span class="comment">INSERT INTO test_data VALUES (6, 'diapers');</span> |
| <a name="l00152"></a>00152 <span class="comment">INSERT INTO test_data VALUES (6, 'chips');</span> |
| <a name="l00153"></a>00153 <span class="comment">INSERT INTO test_data VALUES (7, 'beer');</span> |
| <a name="l00154"></a>00154 <span class="comment">INSERT INTO test_data VALUES (7, 'diapers');</span> |
| <a name="l00155"></a>00155 <span class="comment"></span> |
| <a name="l00156"></a>00156 <span class="comment">\endcode</span> |
| <a name="l00157"></a>00157 <span class="comment"></span> |
| <a name="l00158"></a>00158 <span class="comment">Let \f$ min(support) = .25 \f$ and \f$ min(confidence) = .5 \f$, and the output schema be 'myschema'. For this example, we will set verbose to 'true' so that we have some insight into the progress of the function. We can now generate association rules as follows:</span> |
| <a name="l00159"></a>00159 <span class="comment"></span> |
| <a name="l00160"></a>00160 <span class="comment">\code</span> |
| <a name="l00161"></a>00161 <span class="comment"></span> |
| <a name="l00162"></a>00162 <span class="comment">SELECT * FROM MADLIB_SCHEMA.assoc_rules (.25, .5, 'trans_id', 'product', 'test_data','myschema', false);</span> |
| <a name="l00163"></a>00163 <span class="comment"></span> |
| <a name="l00164"></a>00164 <span class="comment">\endcode</span> |
| <a name="l00165"></a>00165 <span class="comment"></span> |
| <a name="l00166"></a>00166 <span class="comment">This should generate this output:</span> |
| <a name="l00167"></a>00167 <span class="comment"></span> |
| <a name="l00168"></a>00168 <span class="comment">\code</span> |
| <a name="l00169"></a>00169 <span class="comment"></span> |
| <a name="l00170"></a>00170 <span class="comment"> output_schema | output_table | total_rules | total_time </span> |
| <a name="l00171"></a>00171 <span class="comment">---------------+--------------+-------------+-----------------</span> |
| <a name="l00172"></a>00172 <span class="comment"> myschema | assoc_rules | 7 | 00:00:03.162094</span> |
| <a name="l00173"></a>00173 <span class="comment">(1 row)</span> |
| <a name="l00174"></a>00174 <span class="comment"></span> |
| <a name="l00175"></a>00175 <span class="comment"></span> |
| <a name="l00176"></a>00176 <span class="comment">\endcode</span> |
| <a name="l00177"></a>00177 <span class="comment"></span> |
| <a name="l00178"></a>00178 <span class="comment">The association rules are stored in the myschema.assoc_rules:</span> |
| <a name="l00179"></a>00179 <span class="comment"></span> |
| <a name="l00180"></a>00180 <span class="comment">\code</span> |
| <a name="l00181"></a>00181 <span class="comment">select * from myschema.assoc_rules order by support desc;</span> |
| <a name="l00182"></a>00182 <span class="comment"> ruleid | pre | post | support | confidence | lift | conviction </span> |
| <a name="l00183"></a>00183 <span class="comment">--------+-----------------+----------------+-------------------+-------------------+-------------------+-------------------</span> |
| <a name="l00184"></a>00184 <span class="comment"> 4 | {diapers} | {beer} | 0.714285714285714 | 1 | 1 | 0</span> |
| <a name="l00185"></a>00185 <span class="comment"> 2 | {beer} | {diapers} | 0.714285714285714 | 0.714285714285714 | 1 | 1</span> |
| <a name="l00186"></a>00186 <span class="comment"> 1 | {chips} | {beer} | 0.428571428571429 | 1 | 1 | 0</span> |
| <a name="l00187"></a>00187 <span class="comment"> 5 | {chips} | {beer,diapers} | 0.285714285714286 | 0.666666666666667 | 0.933333333333333 | 0.857142857142857</span> |
| <a name="l00188"></a>00188 <span class="comment"> 6 | {chips,beer} | {diapers} | 0.285714285714286 | 0.666666666666667 | 0.933333333333333 | 0.857142857142857</span> |
| <a name="l00189"></a>00189 <span class="comment"> 7 | {chips,diapers} | {beer} | 0.285714285714286 | 1 | 1 | 0</span> |
| <a name="l00190"></a>00190 <span class="comment"> 3 | {chips} | {diapers} | 0.285714285714286 | 0.666666666666667 | 0.933333333333333 | 0.857142857142857</span> |
| <a name="l00191"></a>00191 <span class="comment">(7 rows)</span> |
| <a name="l00192"></a>00192 <span class="comment"></span> |
| <a name="l00193"></a>00193 <span class="comment">\endcode</span> |
| <a name="l00194"></a>00194 <span class="comment"></span> |
| <a name="l00195"></a>00195 <span class="comment">@sa File assoc_rules.sql_in documenting the SQL function.</span> |
| <a name="l00196"></a>00196 <span class="comment"></span> |
| <a name="l00197"></a>00197 <span class="comment">*/</span> |
| <a name="l00198"></a>00198 |
| <a name="l00199"></a>00199 /* |
| <a name="l00200"></a>00200 * @brief The result data type for the association rule API |
| <a name="l00201"></a>00201 * |
| <a name="l00202"></a>00202 * output_schema the name of the output schema. |
| <a name="l00203"></a>00203 * output_table the name of the output table. |
| <a name="l00204"></a>00204 * total_rules the number of total rules. |
| <a name="l00205"></a>00205 * total_time the running time. |
| <a name="l00206"></a>00206 */ |
| <a name="l00207"></a>00207 CREATE TYPE MADLIB_SCHEMA.assoc_rules_results AS |
| <a name="l00208"></a>00208 ( |
| <a name="l00209"></a>00209 output_schema TEXT, |
| <a name="l00210"></a>00210 output_table TEXT, |
| <a name="l00211"></a>00211 total_rules INT, |
| <a name="l00212"></a>00212 total_time INTERVAL |
| <a name="l00213"></a>00213 ); |
| <a name="l00214"></a>00214 |
| <a name="l00215"></a>00215 |
| <a name="l00216"></a>00216 /* |
| <a name="l00217"></a>00217 * @brief Given the text form of a closed frequent pattern (cfp), this function |
| <a name="l00218"></a>00218 * generates the association rules for that pattern. We use text format |
| <a name="l00219"></a>00219 * because text values are hash joinable. The output is a set of text |
| <a name="l00220"></a>00220 * array. For example, assuming the input pattern is '1,2,3<span class="stringliteral">'.</span> |
| <a name="l00221"></a>00221 <span class="stringliteral"> * The result rules:</span> |
| <a name="l00222"></a>00222 <span class="stringliteral"> * array['</span>1<span class="stringliteral">', '</span>2,3<span class="stringliteral">']</span> |
| <a name="l00223"></a>00223 <span class="stringliteral"> * array['</span>2<span class="stringliteral">', '</span>1,3<span class="stringliteral">']</span> |
| <a name="l00224"></a>00224 <span class="stringliteral"> * array['</span>3<span class="stringliteral">', '</span>1,2<span class="stringliteral">']</span> |
| <a name="l00225"></a>00225 <span class="stringliteral"> * array['</span>1,2<span class="stringliteral">', '</span>3<span class="stringliteral">']</span> |
| <a name="l00226"></a>00226 <span class="stringliteral"> * array['</span>1,3<span class="stringliteral">', '</span>2<span class="stringliteral">']</span> |
| <a name="l00227"></a>00227 <span class="stringliteral"> * array['</span>2,3<span class="stringliteral">', '</span>1<span class="stringliteral">']</span> |
| <a name="l00228"></a>00228 <span class="stringliteral"> * Note that two meaningless rules will be excluded:</span> |
| <a name="l00229"></a>00229 <span class="stringliteral"> * array['</span>1,2,3<span class="stringliteral">', NULL]</span> |
| <a name="l00230"></a>00230 <span class="stringliteral"> * array[NULL, '</span>1,2,3<span class="stringliteral">']</span> |
| <a name="l00231"></a>00231 <span class="stringliteral"> *</span> |
| <a name="l00232"></a>00232 <span class="stringliteral"> * @param arg 1 The text form of a closed frequent pattern.</span> |
| <a name="l00233"></a>00233 <span class="stringliteral"> * @param arg 2 The number of items in the pattern.</span> |
| <a name="l00234"></a>00234 <span class="stringliteral"> *</span> |
| <a name="l00235"></a>00235 <span class="stringliteral"> * @return A set of text array. Each array has two elements, corresponding to</span> |
| <a name="l00236"></a>00236 <span class="stringliteral"> * the left and right parts of an association rule.</span> |
| <a name="l00237"></a>00237 <span class="stringliteral"> *</span> |
| <a name="l00238"></a>00238 <span class="stringliteral"> */</span> |
| <a name="l00239"></a>00239 <span class="stringliteral">CREATE OR REPLACE FUNCTION MADLIB_SCHEMA.gen_rules_from_cfp</span> |
| <a name="l00240"></a>00240 <span class="stringliteral"> (</span> |
| <a name="l00241"></a>00241 <span class="stringliteral"> TEXT,</span> |
| <a name="l00242"></a>00242 <span class="stringliteral"> INT</span> |
| <a name="l00243"></a>00243 <span class="stringliteral"> )</span> |
| <a name="l00244"></a>00244 <span class="stringliteral">RETURNS SETOF TEXT[] AS '</span>MODULE_PATHNAME<span class="stringliteral">'</span> |
| <a name="l00245"></a>00245 <span class="stringliteral">LANGUAGE C STRICT IMMUTABLE;</span> |
| <a name="l00246"></a>00246 <span class="stringliteral"></span> |
| <a name="l00247"></a>00247 <span class="stringliteral"></span><span class="comment"></span> |
| <a name="l00248"></a>00248 <span class="comment">/**</span> |
| <a name="l00249"></a>00249 <span class="comment"> *</span> |
| <a name="l00250"></a>00250 <span class="comment"> * @param support minimum level of support needed for each itemset to</span> |
| <a name="l00251"></a>00251 <span class="comment"> * be included in result</span> |
| <a name="l00252"></a>00252 <span class="comment"> * @param confidence minimum level of confidence needed for each rule to</span> |
| <a name="l00253"></a>00253 <span class="comment"> * be included in result</span> |
| <a name="l00254"></a>00254 <span class="comment"> * @param tid_col name of the column storing the transaction ids</span> |
| <a name="l00255"></a>00255 <span class="comment"> * @param item_col name of the column storing the products</span> |
| <a name="l00256"></a>00256 <span class="comment"> * @param input_table name of the table where the data is stored</span> |
| <a name="l00257"></a>00257 <span class="comment"> * @param output_schema name of the schema where the final results will be stored</span> |
| <a name="l00258"></a>00258 <span class="comment"> * @param verbose determining if output contains comments</span> |
| <a name="l00259"></a>00259 <span class="comment"> *</span> |
| <a name="l00260"></a>00260 <span class="comment"> * @returns The schema and table name containing association rules,</span> |
| <a name="l00261"></a>00261 <span class="comment"> * and total number of rules found.</span> |
| <a name="l00262"></a>00262 <span class="comment"> *</span> |
| <a name="l00263"></a>00263 <span class="comment"> * This function computes the association rules between products in a data set.</span> |
| <a name="l00264"></a>00264 <span class="comment"> * It reads the name of the table, the column names of the product and ids, and</span> |
| <a name="l00265"></a>00265 <span class="comment"> * computes ssociation rules using the Apriori algorithm, and subject to the</span> |
| <a name="l00266"></a>00266 <span class="comment"> * support and confidence constraints as input by the user. This version of</span> |
| <a name="l00267"></a>00267 <span class="comment"> * association rules has verbose functionality. When verbose is true, output of</span> |
| <a name="l00268"></a>00268 <span class="comment"> * function includes iteration steps and comments on Apriori algorithm steps.</span> |
| <a name="l00269"></a>00269 <span class="comment"> *</span> |
| <a name="l00270"></a>00270 <span class="comment"> */</span> |
| <a name="l00271"></a>00271 CREATE OR REPLACE FUNCTION MADLIB_SCHEMA.assoc_rules |
| <a name="l00272"></a>00272 ( |
| <a name="l00273"></a>00273 support FLOAT8, |
| <a name="l00274"></a>00274 confidence FLOAT8, |
| <a name="l00275"></a>00275 tid_col TEXT, |
| <a name="l00276"></a>00276 item_col TEXT, |
| <a name="l00277"></a>00277 input_table TEXT, |
| <a name="l00278"></a>00278 output_schema TEXT, |
| <a name="l00279"></a>00279 verbose BOOLEAN |
| <a name="l00280"></a>00280 ) |
| <a name="l00281"></a>00281 RETURNS MADLIB_SCHEMA.assoc_rules_results |
| <a name="l00282"></a>00282 AS $$ |
| <a name="l00283"></a>00283 |
| <a name="l00284"></a>00284 PythonFunctionBodyOnly(`assoc_rules', `<a class="code" href="assoc__rules_8sql__in.html#af9456adb6dad01e452415b9a0a5371dc">assoc_rules</a><span class="stringliteral">')</span> |
| <a name="l00285"></a>00285 <span class="stringliteral"></span> |
| <a name="l00286"></a>00286 <span class="stringliteral"> plpy.execute("SET client_min_messages = error;")</span> |
| <a name="l00287"></a>00287 <span class="stringliteral"></span> |
| <a name="l00288"></a>00288 <span class="stringliteral"> # schema_madlib comes from PythonFunctionBodyOnly</span> |
| <a name="l00289"></a>00289 <span class="stringliteral"> return assoc_rules.assoc_rules(</span> |
| <a name="l00290"></a>00290 <span class="stringliteral"> schema_madlib,</span> |
| <a name="l00291"></a>00291 <span class="stringliteral"> support,</span> |
| <a name="l00292"></a>00292 <span class="stringliteral"> confidence,</span> |
| <a name="l00293"></a><a class="code" href="assoc__rules_8sql__in.html#af9456adb6dad01e452415b9a0a5371dc">00293</a> <span class="stringliteral"> tid_col,</span> |
| <a name="l00294"></a>00294 <span class="stringliteral"> item_col,</span> |
| <a name="l00295"></a>00295 <span class="stringliteral"> input_table,</span> |
| <a name="l00296"></a>00296 <span class="stringliteral"> output_schema,</span> |
| <a name="l00297"></a>00297 <span class="stringliteral"> verbose</span> |
| <a name="l00298"></a>00298 <span class="stringliteral"> );</span> |
| <a name="l00299"></a>00299 <span class="stringliteral"></span> |
| <a name="l00300"></a>00300 <span class="stringliteral">$$ LANGUAGE plpythonu;</span> |
| <a name="l00301"></a>00301 <span class="stringliteral"></span> |
| <a name="l00302"></a>00302 <span class="stringliteral"></span><span class="comment"></span> |
| <a name="l00303"></a>00303 <span class="comment">/**</span> |
| <a name="l00304"></a>00304 <span class="comment"> *</span> |
| <a name="l00305"></a>00305 <span class="comment"> * @brief The short form of the above function with vobose removed.</span> |
| <a name="l00306"></a>00306 <span class="comment"> *</span> |
| <a name="l00307"></a>00307 <span class="comment"> */</span> |
| <a name="l00308"></a>00308 CREATE OR REPLACE FUNCTION MADLIB_SCHEMA.assoc_rules |
| <a name="l00309"></a>00309 ( |
| <a name="l00310"></a>00310 support FLOAT8, |
| <a name="l00311"></a>00311 confidence FLOAT8, |
| <a name="l00312"></a>00312 tid_col TEXT, |
| <a name="l00313"></a>00313 item_col TEXT, |
| <a name="l00314"></a>00314 input_table TEXT, |
| <a name="l00315"></a>00315 output_schema TEXT |
| <a name="l00316"></a>00316 ) |
| <a name="l00317"></a>00317 RETURNS MADLIB_SCHEMA.assoc_rules_results |
| <a name="l00318"></a>00318 AS $$ |
| <a name="l00319"></a>00319 |
| <a name="l00320"></a>00320 PythonFunctionBodyOnly(`assoc_rules', `<a class="code" href="assoc__rules_8sql__in.html#af9456adb6dad01e452415b9a0a5371dc">assoc_rules</a><span class="stringliteral">')</span> |
| <a name="l00321"></a>00321 <span class="stringliteral"></span> |
| <a name="l00322"></a>00322 <span class="stringliteral"> plpy.execute("SET client_min_messages = error;")</span> |
| <a name="l00323"></a>00323 <span class="stringliteral"></span> |
| <a name="l00324"></a>00324 <span class="stringliteral"> # schema_madlib comes from PythonFunctionBodyOnly</span> |
| <a name="l00325"></a>00325 <span class="stringliteral"> return assoc_rules.assoc_rules(</span> |
| <a name="l00326"></a>00326 <span class="stringliteral"> schema_madlib,</span> |
| <a name="l00327"></a>00327 <span class="stringliteral"> support,</span> |
| <a name="l00328"></a>00328 <span class="stringliteral"> confidence,</span> |
| <a name="l00329"></a>00329 <span class="stringliteral"> tid_col,</span> |
| <a name="l00330"></a><a class="code" href="assoc__rules_8sql__in.html#a68a256d98b82ac15bac7df92e806f6f8">00330</a> <span class="stringliteral"> item_col,</span> |
| <a name="l00331"></a>00331 <span class="stringliteral"> input_table,</span> |
| <a name="l00332"></a>00332 <span class="stringliteral"> output_schema,</span> |
| <a name="l00333"></a>00333 <span class="stringliteral"> False</span> |
| <a name="l00334"></a>00334 <span class="stringliteral"> );</span> |
| <a name="l00335"></a>00335 <span class="stringliteral"></span> |
| <a name="l00336"></a>00336 <span class="stringliteral">$$ LANGUAGE plpythonu;</span> |
| </pre></div></div> |
| </div> |
| <div id="nav-path" class="navpath"> |
| <ul> |
| <li class="navelem"><a class="el" href="assoc__rules_8sql__in.html">assoc_rules.sql_in</a> </li> |
| <!-- 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></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> |
| |
| |
| <li class="footer">Generated on Tue Apr 2 2013 14:57:03 for MADlib by |
| <a href="http://www.doxygen.org/index.html"> |
| <img class="footer" src="doxygen.png" alt="doxygen"/></a> 1.7.5.1 </li> |
| </ul> |
| </div> |
| |
| |
| </body> |
| </html> |
