You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@madlib.apache.org by xt...@apache.org on 2016/09/20 18:31:22 UTC
[10/51] [partial] incubator-madlib-site git commit: Update doc for
1.9.1 release
http://git-wip-us.apache.org/repos/asf/incubator-madlib-site/blob/bed9253d/docs/v1.9.1/graph_legend.svg
----------------------------------------------------------------------
diff --git a/docs/v1.9.1/graph_legend.svg b/docs/v1.9.1/graph_legend.svg
new file mode 100644
index 0000000..12d6a4d
--- /dev/null
+++ b/docs/v1.9.1/graph_legend.svg
@@ -0,0 +1,121 @@
+<?xml version="1.0" encoding="UTF-8" standalone="no"?>
+<!DOCTYPE svg PUBLIC "-//W3C//DTD SVG 1.1//EN"
+ "http://www.w3.org/Graphics/SVG/1.1/DTD/svg11.dtd">
+<!-- Generated by graphviz version 2.38.0 (20140413.2041)
+ -->
+<!-- Title: Graph Legend Pages: 1 -->
+<svg width="510pt" height="154pt"
+ viewBox="0.00 0.00 509.69 154.00" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink">
+<g id="graph0" class="graph" transform="scale(1 1) rotate(0) translate(4 150)">
+<title>Graph Legend</title>
+<polygon fill="white" stroke="none" points="-4,4 -4,-150 505.687,-150 505.687,4 -4,4"/>
+<!-- Node9 -->
+<g id="node1" class="node"><title>Node9</title>
+<polygon fill="#bfbfbf" stroke="black" points="293.431,-18 238.599,-18 238.599,-0 293.431,-0 293.431,-18"/>
+<text text-anchor="middle" x="266.015" y="-6" font-family="Helvetica,sans-Serif" font-size="10.00">Inherited</text>
+</g>
+<!-- Node10 -->
+<g id="node2" class="node"><title>Node10</title>
+<g id="a_node2"><a xlink:href="$classPublicBase.html" xlink:title="PublicBase">
+<polygon fill="none" stroke="black" points="66.044,-82 -0.0146549,-82 -0.0146549,-64 66.044,-64 66.044,-82"/>
+<text text-anchor="middle" x="33.0146" y="-70" font-family="Helvetica,sans-Serif" font-size="10.00">PublicBase</text>
+</a>
+</g>
+</g>
+<!-- Node10->Node9 -->
+<g id="edge1" class="edge"><title>Node10->Node9</title>
+<path fill="none" stroke="midnightblue" d="M73.3674,-61.2624C120.605,-48.6928 197.564,-28.2142 238.451,-17.3346"/>
+<polygon fill="midnightblue" stroke="midnightblue" points="72.239,-57.9407 63.4753,-63.8946 74.0391,-64.7054 72.239,-57.9407"/>
+</g>
+<!-- Node11 -->
+<g id="node3" class="node"><title>Node11</title>
+<g id="a_node3"><a xlink:href="$classTruncated.html" xlink:title="Truncated">
+<polygon fill="none" stroke="red" points="63.1738,-146 2.85549,-146 2.85549,-128 63.1738,-128 63.1738,-146"/>
+<text text-anchor="middle" x="33.0146" y="-134" font-family="Helvetica,sans-Serif" font-size="10.00">Truncated</text>
+</a>
+</g>
+</g>
+<!-- Node11->Node10 -->
+<g id="edge2" class="edge"><title>Node11->Node10</title>
+<path fill="none" stroke="midnightblue" d="M33.0146,-117.828C33.0146,-106.079 33.0146,-91.3449 33.0146,-82.3053"/>
+<polygon fill="midnightblue" stroke="midnightblue" points="29.5147,-117.895 33.0146,-127.895 36.5147,-117.895 29.5147,-117.895"/>
+</g>
+<!-- Node13 -->
+<g id="node4" class="node"><title>Node13</title>
+<g id="a_node4"><a xlink:href="$classProtectedBase.html" xlink:title="ProtectedBase">
+<polygon fill="none" stroke="black" points="165.611,-82 84.4179,-82 84.4179,-64 165.611,-64 165.611,-82"/>
+<text text-anchor="middle" x="125.015" y="-70" font-family="Helvetica,sans-Serif" font-size="10.00">ProtectedBase</text>
+</a>
+</g>
+</g>
+<!-- Node13->Node9 -->
+<g id="edge3" class="edge"><title>Node13->Node9</title>
+<path fill="none" stroke="darkgreen" d="M152.713,-59.8204C180.778,-47.4799 223.402,-28.7375 247.577,-18.1074"/>
+<polygon fill="darkgreen" stroke="darkgreen" points="151.193,-56.6654 143.448,-63.8946 154.011,-63.0733 151.193,-56.6654"/>
+</g>
+<!-- Node14 -->
+<g id="node5" class="node"><title>Node14</title>
+<g id="a_node5"><a xlink:href="$classPrivateBase.html" xlink:title="PrivateBase">
+<polygon fill="none" stroke="black" points="253.931,-82 184.099,-82 184.099,-64 253.931,-64 253.931,-82"/>
+<text text-anchor="middle" x="219.015" y="-70" font-family="Helvetica,sans-Serif" font-size="10.00">PrivateBase</text>
+</a>
+</g>
+</g>
+<!-- Node14->Node9 -->
+<g id="edge4" class="edge"><title>Node14->Node9</title>
+<path fill="none" stroke="#8b1a1a" d="M231.432,-55.6198C240.534,-43.6129 252.513,-27.8108 259.719,-18.3053"/>
+<polygon fill="#8b1a1a" stroke="#8b1a1a" points="228.411,-53.8111 225.159,-63.8946 233.989,-58.0399 228.411,-53.8111"/>
+</g>
+<!-- Node15 -->
+<g id="node6" class="node"><title>Node15</title>
+<polygon fill="none" stroke="#bfbfbf" points="355.837,-82 272.192,-82 272.192,-64 355.837,-64 355.837,-82"/>
+<text text-anchor="middle" x="314.015" y="-70" font-family="Helvetica,sans-Serif" font-size="10.00">Undocumented</text>
+</g>
+<!-- Node15->Node9 -->
+<g id="edge5" class="edge"><title>Node15->Node9</title>
+<path fill="none" stroke="midnightblue" d="M301.606,-55.9718C292.278,-43.9234 279.876,-27.904 272.445,-18.3053"/>
+<polygon fill="midnightblue" stroke="midnightblue" points="298.85,-58.1299 307.739,-63.8946 304.385,-53.8447 298.85,-58.1299"/>
+</g>
+<!-- Node16 -->
+<g id="node7" class="node"><title>Node16</title>
+<g id="a_node7"><a xlink:href="$classTempl.html" xlink:title="Templ< int >">
+<polygon fill="none" stroke="black" points="444.489,-82 373.54,-82 373.54,-64 444.489,-64 444.489,-82"/>
+<text text-anchor="middle" x="409.015" y="-70" font-family="Helvetica,sans-Serif" font-size="10.00">Templ< int ></text>
+</a>
+</g>
+</g>
+<!-- Node16->Node9 -->
+<g id="edge6" class="edge"><title>Node16->Node9</title>
+<path fill="none" stroke="midnightblue" d="M380.923,-59.8204C352.46,-47.4799 309.232,-28.7375 284.714,-18.1074"/>
+<polygon fill="midnightblue" stroke="midnightblue" points="379.753,-63.1278 390.32,-63.8946 382.537,-56.7055 379.753,-63.1278"/>
+</g>
+<!-- Node17 -->
+<g id="node8" class="node"><title>Node17</title>
+<g id="a_node8"><a xlink:href="$classTempl.html" xlink:title="Templ< T >">
+<polygon fill="none" stroke="black" points="441.683,-146 376.346,-146 376.346,-128 441.683,-128 441.683,-146"/>
+<text text-anchor="middle" x="409.015" y="-134" font-family="Helvetica,sans-Serif" font-size="10.00">Templ< T ></text>
+</a>
+</g>
+</g>
+<!-- Node17->Node16 -->
+<g id="edge7" class="edge"><title>Node17->Node16</title>
+<path fill="none" stroke="orange" stroke-dasharray="5,2" d="M409.015,-117.828C409.015,-106.079 409.015,-91.3449 409.015,-82.3053"/>
+<polygon fill="orange" stroke="orange" points="405.515,-117.895 409.015,-127.895 412.515,-117.895 405.515,-117.895"/>
+<text text-anchor="middle" x="422.914" y="-102" font-family="Helvetica,sans-Serif" font-size="10.00">< int ></text>
+</g>
+<!-- Node18 -->
+<g id="node9" class="node"><title>Node18</title>
+<g id="a_node9"><a xlink:href="$classUsed.html" xlink:title="Used">
+<polygon fill="none" stroke="black" points="501.861,-82 462.168,-82 462.168,-64 501.861,-64 501.861,-82"/>
+<text text-anchor="middle" x="482.015" y="-70" font-family="Helvetica,sans-Serif" font-size="10.00">Used</text>
+</a>
+</g>
+</g>
+<!-- Node18->Node9 -->
+<g id="edge8" class="edge"><title>Node18->Node9</title>
+<path fill="none" stroke="#9a32cd" stroke-dasharray="5,2" d="M452.451,-63.5142C410.543,-51.4849 334.429,-29.6374 293.618,-17.9232"/>
+<polygon fill="#9a32cd" stroke="#9a32cd" points="451.504,-66.8834 462.081,-66.2783 453.435,-60.1551 451.504,-66.8834"/>
+<text text-anchor="middle" x="417.305" y="-38" font-family="Helvetica,sans-Serif" font-size="10.00">m_usedClass</text>
+</g>
+</g>
+</svg>
http://git-wip-us.apache.org/repos/asf/incubator-madlib-site/blob/bed9253d/docs/v1.9.1/group__grp__arima.html
----------------------------------------------------------------------
diff --git a/docs/v1.9.1/group__grp__arima.html b/docs/v1.9.1/group__grp__arima.html
new file mode 100644
index 0000000..341d112
--- /dev/null
+++ b/docs/v1.9.1/group__grp__arima.html
@@ -0,0 +1,415 @@
+<!-- HTML header for doxygen 1.8.4-->
+<!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 1.8.10"/>
+<meta name="keywords" content="madlib,postgres,greenplum,machine learning,data mining,deep learning,ensemble methods,data science,market basket analysis,affinity analysis,pca,lda,regression,elastic net,huber white,proportional hazards,k-means,latent dirichlet allocation,bayes,support vector machines,svm"/>
+<title>MADlib: ARIMA</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="navtreedata.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/searchdata.js"></script>
+<script type="text/javascript" src="search/search.js"></script>
+<script type="text/javascript">
+ $(document).ready(function() { init_search(); });
+</script>
+<!-- hack in the navigation tree -->
+<script type="text/javascript" src="eigen_navtree_hacks.js"></script>
+<link href="doxygen.css" rel="stylesheet" type="text/css" />
+<link href="madlib_extra.css" rel="stylesheet" type="text/css"/>
+<!-- google analytics -->
+<script>
+ (function(i,s,o,g,r,a,m){i['GoogleAnalyticsObject']=r;i[r]=i[r]||function(){
+ (i[r].q=i[r].q||[]).push(arguments)},i[r].l=1*new Date();a=s.createElement(o),
+ m=s.getElementsByTagName(o)[0];a.async=1;a.src=g;m.parentNode.insertBefore(a,m)
+ })(window,document,'script','//www.google-analytics.com/analytics.js','ga');
+ ga('create', 'UA-45382226-1', 'madlib.net');
+ ga('send', 'pageview');
+</script>
+</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 id="projectlogo"><a href="http://madlib.net"><img alt="Logo" src="madlib.png" height="50" style="padding-left:0.5em;" border="0"/ ></a></td>
+ <td style="padding-left: 0.5em;">
+ <div id="projectname">
+ <span id="projectnumber">1.9.1</span>
+ </div>
+ <div id="projectbrief">User Documentation for MADlib</div>
+ </td>
+ <td> <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>
+</td>
+ </tr>
+ </tbody>
+</table>
+</div>
+<!-- end header part -->
+<!-- Generated by Doxygen 1.8.10 -->
+<script type="text/javascript">
+var searchBox = new SearchBox("searchBox", "search",false,'Search');
+</script>
+</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('group__grp__arima.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)">
+</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">ARIMA<div class="ingroups"><a class="el" href="group__grp__tsa.html">Time Series Analysis</a></div></div> </div>
+</div><!--header-->
+<div class="contents">
+<div class="toc"><b>Contents</b> </p><ul>
+<li class="level1">
+<a href="#train">Training Function</a> </li>
+<li class="level1">
+<a href="#forecast">Forecasting Function</a> </li>
+<li class="level1">
+<a href="#examples">Examples</a> </li>
+<li class="level1">
+<a href="#background">Technical Background</a> </li>
+<li class="level1">
+<a href="#literature">Literature</a> </li>
+<li class="level1">
+<a href="#related">Related Topics</a> </li>
+</ul>
+</div><p>Given a time series of data X, the Autoregressive Integrated Moving Average (ARIMA) model is a tool for understanding and, perhaps, predicting future values in the series. The model consists of three parts, an autoregressive (AR) part, a moving average (MA) part, and an integrated (I) part where an initial differencing step can be applied to remove any non-stationarity in the signal. The model is generally referred to as an ARIMA(p, d, q) model where parameters p, d, and q are non-negative integers that refer to the order of the autoregressive, integrated, and moving average parts of the model respectively.</p>
+<p><a class="anchor" id="train"></a></p><dl class="section user"><dt>Training Function</dt><dd></dd></dl>
+<p>The ARIMA training function has the following syntax. </p><pre class="syntax">
+arima_train( input_table,
+ output_table,
+ timestamp_column,
+ timeseries_column,
+ grouping_columns,
+ include_mean,
+ non_seasonal_orders,
+ optimizer_params
+ )
+</pre><p><b>Arguments</b> </p><dl class="arglist">
+<dt>input_table </dt>
+<dd><p class="startdd">TEXT. The name of the table containing time series data.</p>
+<p class="enddd"></p>
+</dd>
+<dt>output_table </dt>
+<dd><p class="startdd">TEXT. The name of the table to store the ARIMA model. Three tables are created, with names based on the value of the <em>output_table</em> argument in the training function:</p>
+<ol type="1">
+<li><em>output_table</em>: Table containing the ARIMA model. Contains the following columns: <table class="output">
+<tr>
+<th>mean </th><td>Model mean (only if 'include_mean' is TRUE) </td></tr>
+<tr>
+<th>mean_std_error </th><td>Standard errors for mean </td></tr>
+<tr>
+<th>ar_params </th><td>Auto-regressions parameters of the ARIMA model </td></tr>
+<tr>
+<th>ar_std_errors </th><td>Standard errors for AR parameters </td></tr>
+<tr>
+<th>ma_params </th><td>Moving average parameters of the ARIMA model </td></tr>
+<tr>
+<th>ma_std_errors </th><td>Standard errors for MA parameters </td></tr>
+</table>
+</li>
+<li><em>output_table</em>_summary: Table containing descriptive statistics of the ARIMA model. Contains the following columns: <table class="output">
+<tr>
+<th>input_table </th><td>Table name with the source data </td></tr>
+<tr>
+<th>timestamp_col </th><td>Column name in the source table that contains the timestamp index to data </td></tr>
+<tr>
+<th>timeseries_col </th><td>Column name in the source table that contains the data values </td></tr>
+<tr>
+<th>non_seasonal_orders </th><td>Orders of the non-seasonal ARIMA model </td></tr>
+<tr>
+<th>include_mean </th><td>TRUE if intercept was included in ARIMA model </td></tr>
+<tr>
+<th>residual_variance </th><td>Variance of the residuals </td></tr>
+<tr>
+<th>log_likelihood </th><td>Log likelihood value (when using MLE) </td></tr>
+<tr>
+<th>iter_num </th><td>The number of iterations executed </td></tr>
+<tr>
+<th>exec_time </th><td>Total time taken to train the model </td></tr>
+</table>
+</li>
+<li><em>output_table</em>_residual: Table containing the residuals for each data point in 'input_table'. Contains the following columns: <table class="output">
+<tr>
+<th>timestamp_col </th><td>Same as the 'timestamp_col' parameter (all indices from source table included except the first <em>d</em> elements, where <em>d</em> is the differencing order value from 'non_seasonal_orders') </td></tr>
+<tr>
+<th>residual </th><td>Residual value for each data point </td></tr>
+</table>
+</li>
+</ol>
+<p></p>
+<p class="enddd"></p>
+</dd>
+<dt>timestamp_column </dt>
+<dd><p class="startdd">TEXT. The name of the column containing the timestamp (index) data. This could be a serial index (INTEGER) or date/time value (TIMESTAMP).</p>
+<p class="enddd"></p>
+</dd>
+<dt>timeseries_column </dt>
+<dd><p class="startdd">TEXT. The name of the column containing the time series data. This data is currently restricted to DOUBLE PRECISION.</p>
+<p class="enddd"></p>
+</dd>
+<dt>grouping_columns (optional) </dt>
+<dd><p class="startdd">TEXT, default: NULL. <em>Not currently implemented. Any non-NULL value is ignored.</em></p>
+<p>A comma-separated list of column names used to group the input dataset into discrete groups, training one ARIMA model per group. It is similar to the SQL <code>GROUP BY</code> clause. When this value is null, no grouping is used and a single result model is generated.</p>
+<p class="enddd"></p>
+</dd>
+<dt>include_mean (optional) </dt>
+<dd><p class="startdd">BOOLEAN, default: FALSE. Mean value of the data series is added in the ARIMA model if this variable is True. </p>
+<p class="enddd"></p>
+</dd>
+<dt>non_seasonal_orders (optional) </dt>
+<dd><p class="startdd">INTEGER[], default: 'ARRAY[1,1,1]'. Orders of the ARIMA model. The orders are [p, d, q], where parameters p, d, and q are non-negative integers that refer to the order of the autoregressive, integrated, and moving average parts of the model respectively. </p>
+<p class="enddd"></p>
+</dd>
+<dt>optimizer_params (optional) </dt>
+<dd>TEXT. Comma-separated list of optimizer-specific parameters of the form \u2018name=value'. The order of the parameters does not matter. The following parameters are recognized:<ul>
+<li><b>max_iter:</b> Maximum number of iterations to run learning algorithm (Default = 100)</li>
+<li><b>tau:</b> Computes the initial step size for gradient algorithm (Default = 0.001)</li>
+<li><b>e1:</b> Algorithm-specific threshold for convergence (Default = 1e-15)</li>
+<li><b>e2:</b> Algorithm-specific threshold for convergence (Default = 1e-15)</li>
+<li><b>e3:</b> Algorithm-specific threshold for convergence (Default = 1e-15)</li>
+<li><b>hessian_delta:</b> Delta parameter to compute a numerical approximation of the Hessian matrix (Default = 1e-6) </li>
+</ul>
+</dd>
+</dl>
+<p><a class="anchor" id="forecast"></a></p><dl class="section user"><dt>Forecasting Function</dt><dd></dd></dl>
+<p>The ARIMA forecast function has the following syntax. </p><pre class="syntax">
+arima_forecast( model_table,
+ output_table,
+ steps_ahead
+ )
+</pre><p> <b>Arguments</b> </p><dl class="arglist">
+<dt>model_table </dt>
+<dd><p class="startdd">TEXT. The name of the table containing the ARIMA model trained on the time series dataset.</p>
+<p class="enddd"></p>
+</dd>
+<dt>output_table </dt>
+<dd><p class="startdd">TEXT. The name of the table to store the forecasted values. The output table produced by the forecast function contains the following columns. </p><table class="output">
+<tr>
+<th>group_by_cols </th><td>Grouping column values (if grouping parameter is provided) </td></tr>
+<tr>
+<th>step_ahead </th><td>Time step for the forecast </td></tr>
+<tr>
+<th>forecast_value </th><td>Forecast of the current time step </td></tr>
+</table>
+<p class="enddd"></p>
+</dd>
+<dt>steps_ahead </dt>
+<dd>INTEGER. The number of steps to forecast at the end of the time series. </dd>
+</dl>
+<p><a class="anchor" id="examples"></a></p><dl class="section user"><dt>Examples</dt><dd><ol type="1">
+<li>View online help for the ARIMA training function. <pre class="example">
+SELECT madlib.arima_train();
+</pre></li>
+<li>Create an input data set. <pre class="example">
+DROP TABLE IF EXISTS arima_beer;
+CREATE TABLE arima_beer (time_id integer NOT NULL, value double precision NOT NULL );
+COPY arima_beer (time_id, value) FROM stdin WITH DELIMITER '|';
+1 | 93.2
+2 | 96.0
+3 | 95.2
+4 | 77.0
+5 | 70.9
+6 | 64.7
+7 | 70.0
+8 | 77.2
+9 | 79.5
+10 | 100.5
+11 | 100.7
+12 | 107.0
+13 | 95.9
+14 | 82.7
+15 | 83.2
+16 | 80.0
+17 | 80.4
+18 | 67.5
+19 | 75.7
+20 | 71.0
+21 | 89.2
+22 | 101.0
+23 | 105.2
+24 | 114.0
+25 | 96.2
+26 | 84.4
+27 | 91.2
+28 | 81.9
+29 | 80.5
+30 | 70.4
+31 | 74.7
+32 | 75.9
+33 | 86.2
+34 | 98.7
+35 | 100.9
+36 | 113.7
+37 | 89.7
+38 | 84.4
+39 | 87.2
+40 | 85.5
+\.
+</pre></li>
+<li>Train an ARIMA model. <pre class="example">
+-- Train ARIMA model with 'grouping_columns'=NULL, 'include_mean'=TRUE,
+-- and 'non_seasonal_orders'=[1,1,1]
+SELECT madlib.arima_train( 'arima_beer',
+ 'arima_beer_output',
+ 'time_id',
+ 'value',
+ NULL,
+ FALSE,
+ ARRAY[1, 1, 1]
+ );
+</pre></li>
+<li>Examine the ARIMA model. <pre class="example">
+\x ON
+SELECT * FROM arima_beer_output;
+</pre> Result: <pre class="result">
+-[ RECORD 1 ]-+------------------
+ar_params | {0.221954769696}
+ar_std_errors | {0.575367782602}
+ma_params | {-0.140623564576}
+ma_std_errors | {0.533445214346}
+</pre></li>
+<li>View the summary statistics table. <pre class="example">
+SELECT * FROM arima_beer_output_summary;
+</pre> Result: <pre class="result">
+-[ RECORD 1 ]-------+---------------
+input_table | arima_beer
+timestamp_col | time_id
+timeseries_col | value
+non_seasonal_orders | {1,1,1}
+include_mean | f
+residual_variance | 100.989970539
+log_likelihood | -145.331516396
+iter_num | 28
+exec_time (s) | 2.75
+</pre></li>
+<li>View the residuals. <pre class="example">
+\x OFF
+SELECT * FROM arima_beer_output_residual;
+</pre> Result: <pre class="result">
+ time_id | residual
+---------+--------------------
+ 2 | 0
+ 4 | -18.222328834394
+ 6 | -5.49616627282665
+...
+ 35 | 1.06298837051437
+ 37 | -25.0886854003757
+ 39 | 3.48401666299571
+(40 rows)
+</pre></li>
+<li>Use the ARIMA forecast function to forecast 10 future values. <pre class="example">
+SELECT madlib.arima_forecast( 'arima_beer_output',
+ 'arima_beer_forecast_output',
+ 10
+ );
+SELECT * FROM arima_beer_forecast_output;
+</pre> Result: <pre class="result">
+ steps_ahead | forecast_value
+-------------+----------------
+ 1 | 85.3802343659
+ 3 | 85.3477516875
+ 5 | 85.3461514635
+ 7 | 85.3460726302
+ 9 | 85.3460687465
+ 2 | 85.3536518121
+ 4 | 85.3464421267
+ 6 | 85.3460869494
+ 8 | 85.3460694519
+ 10 | 85.34606859
+(10 rows)
+</pre></li>
+</ol>
+</dd></dl>
+<p><a class="anchor" id="background"></a></p><dl class="section user"><dt>Technical Background</dt><dd>An ARIMA model is an <em>a</em>uto-<em>r</em>egressive <em>i</em>ntegrated <em>m</em>oving <em>a</em>verage model. An ARIMA model is typically expressed in the form <p class="formulaDsp">
+<img class="formulaDsp" alt="\[ (1 - \phi(B)) Y_t = (1 + \theta(B)) Z_t, \]" src="form_529.png"/>
+</p>
+</dd></dl>
+<p>where <img class="formulaInl" alt="$B$" src="form_206.png"/> is the backshift operator. The time <img class="formulaInl" alt="$ t $" src="form_530.png"/> is from <img class="formulaInl" alt="$ 1 $" src="form_531.png"/> to <img class="formulaInl" alt="$ N $" src="form_219.png"/>.</p>
+<p>ARIMA models involve the following variables:</p><ul>
+<li>The values of the time series: <img class="formulaInl" alt="$ X_t $" src="form_532.png"/>.</li>
+<li>Parameters of the model: <img class="formulaInl" alt="$ p $" src="form_110.png"/>, <img class="formulaInl" alt="$ q $" src="form_533.png"/>, and <img class="formulaInl" alt="$ d $" src="form_468.png"/>; <img class="formulaInl" alt="$ d $" src="form_468.png"/> is the differencing order, <img class="formulaInl" alt="$ p $" src="form_110.png"/> is the order of the AR operator, and <img class="formulaInl" alt="$ q $" src="form_533.png"/> is the order of the MA operator.</li>
+<li>The AR operator: <img class="formulaInl" alt="$ \phi(B) $" src="form_534.png"/>.</li>
+<li>The MA operator: <img class="formulaInl" alt="$ \theta(B) $" src="form_535.png"/>.</li>
+<li>The lag difference: <img class="formulaInl" alt="$ Y_{t} $" src="form_536.png"/>, where <img class="formulaInl" alt="$ Y_{t} = (1-B)^{d}(X_{t} - \mu) $" src="form_537.png"/>.</li>
+<li>The mean value: <img class="formulaInl" alt="$ \mu $" src="form_287.png"/>, which is set to be zero for <img class="formulaInl" alt="$ d>0 $" src="form_538.png"/> and estimated from the data when d=0.</li>
+<li>The error terms: <img class="formulaInl" alt="$ Z_t $" src="form_539.png"/>.</li>
+</ul>
+<p>The auto regression operator models the prediction for the next observation as some linear combination of the previous observations. More formally, an AR operator of order <img class="formulaInl" alt="$ p $" src="form_110.png"/> is defined as</p>
+<p class="formulaDsp">
+<img class="formulaDsp" alt="\[ \phi(B) Y_t= \phi_1 Y_{t-1} + \dots + \phi_{p} Y_{t-p} \]" src="form_540.png"/>
+</p>
+<p>The moving average operator is similar, and it models the prediction for the next observation as a linear combination of the errors in the previous prediction errors. More formally, the MA operator of order <img class="formulaInl" alt="$ q $" src="form_533.png"/> is defined as</p>
+<p class="formulaDsp">
+<img class="formulaDsp" alt="\[ \theta(B) Z_t = \theta_{1} Z_{t-1} + \dots + \theta_{q} Z_{t-q}. \]" src="form_541.png"/>
+</p>
+<p>We estimate the parameters using the Levenberg-Marquardt Algorithm. In mathematics and computing, the Levenberg-Marquardt algorithm (LMA), also known as the damped least-squares (DLS) method, provides a numerical solution to the problem of minimizing a function, generally nonlinear, over a space of parameters of the function.</p>
+<p>Like other numeric minimization algorithms, LMA is an iterative procedure. To start a minimization, the user has to provide an initial guess for the parameter vector, $p$, as well as some tuning parameters <img class="formulaInl" alt="$\tau, \epsilon_1, \epsilon_2, \epsilon_3,$" src="form_542.png"/>.</p>
+<p><a class="anchor" id="literature"></a></p><dl class="section user"><dt>Literature</dt><dd></dd></dl>
+<p>[1] Rob J Hyndman and George Athanasopoulos: Forecasting: principles and practice, <a href="http://otexts.com/fpp/">http://otexts.com/fpp/</a></p>
+<p>[2] Robert H. Shumway, David S. Stoffer: Time Series Analysis and Its Applications With R Examples, Third edition Springer Texts in Statistics, 2010</p>
+<p>[3] Henri Gavin: The Levenberg-Marquardt method for nonlinear least squares curve-fitting problems, 2011</p>
+<p><a class="anchor" id="related"></a></p><dl class="section user"><dt>Related Topics</dt><dd></dd></dl>
+<p>File <a class="el" href="arima_8sql__in.html" title="Arima function for forecasting of timeseries data. ">arima.sql_in</a> documenting the ARIMA functions </p>
+</div><!-- contents -->
+</div><!-- doc-content -->
+<!-- start footer part -->
+<div id="nav-path" class="navpath"><!-- id is needed for treeview function! -->
+ <ul>
+ <li class="footer">Generated on Tue Sep 20 2016 11:27:01 for MADlib by
+ <a href="http://www.doxygen.org/index.html">
+ <img class="footer" src="doxygen.png" alt="doxygen"/></a> 1.8.10 </li>
+ </ul>
+</div>
+</body>
+</html>
http://git-wip-us.apache.org/repos/asf/incubator-madlib-site/blob/bed9253d/docs/v1.9.1/group__grp__array.html
----------------------------------------------------------------------
diff --git a/docs/v1.9.1/group__grp__array.html b/docs/v1.9.1/group__grp__array.html
new file mode 100644
index 0000000..ed276e6
--- /dev/null
+++ b/docs/v1.9.1/group__grp__array.html
@@ -0,0 +1,328 @@
+<!-- HTML header for doxygen 1.8.4-->
+<!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 1.8.10"/>
+<meta name="keywords" content="madlib,postgres,greenplum,machine learning,data mining,deep learning,ensemble methods,data science,market basket analysis,affinity analysis,pca,lda,regression,elastic net,huber white,proportional hazards,k-means,latent dirichlet allocation,bayes,support vector machines,svm"/>
+<title>MADlib: Array Operations</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="navtreedata.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/searchdata.js"></script>
+<script type="text/javascript" src="search/search.js"></script>
+<script type="text/javascript">
+ $(document).ready(function() { init_search(); });
+</script>
+<!-- hack in the navigation tree -->
+<script type="text/javascript" src="eigen_navtree_hacks.js"></script>
+<link href="doxygen.css" rel="stylesheet" type="text/css" />
+<link href="madlib_extra.css" rel="stylesheet" type="text/css"/>
+<!-- google analytics -->
+<script>
+ (function(i,s,o,g,r,a,m){i['GoogleAnalyticsObject']=r;i[r]=i[r]||function(){
+ (i[r].q=i[r].q||[]).push(arguments)},i[r].l=1*new Date();a=s.createElement(o),
+ m=s.getElementsByTagName(o)[0];a.async=1;a.src=g;m.parentNode.insertBefore(a,m)
+ })(window,document,'script','//www.google-analytics.com/analytics.js','ga');
+ ga('create', 'UA-45382226-1', 'madlib.net');
+ ga('send', 'pageview');
+</script>
+</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 id="projectlogo"><a href="http://madlib.net"><img alt="Logo" src="madlib.png" height="50" style="padding-left:0.5em;" border="0"/ ></a></td>
+ <td style="padding-left: 0.5em;">
+ <div id="projectname">
+ <span id="projectnumber">1.9.1</span>
+ </div>
+ <div id="projectbrief">User Documentation for MADlib</div>
+ </td>
+ <td> <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>
+</td>
+ </tr>
+ </tbody>
+</table>
+</div>
+<!-- end header part -->
+<!-- Generated by Doxygen 1.8.10 -->
+<script type="text/javascript">
+var searchBox = new SearchBox("searchBox", "search",false,'Search');
+</script>
+</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('group__grp__array.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)">
+</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">Array Operations<div class="ingroups"><a class="el" href="group__grp__datatrans.html">Data Types and Transformations</a> » <a class="el" href="group__grp__arraysmatrix.html">Arrays and Matrices</a></div></div> </div>
+</div><!--header-->
+<div class="contents">
+<div class="toc"><b>Contents</b> </p><ul>
+<li>
+<a href="#notes">Implementation Notes</a> </li>
+<li>
+<a href="#list">List of Array Operations</a> </li>
+<li>
+<a href="#examples">Examples</a> </li>
+<li>
+<a href="#related">Related Topics</a> </li>
+</ul>
+</div><p>This module provides a set of basic array operations implemented in C. It is a support module for several machine learning algorithms that require fast array operations.</p>
+<p><a class="anchor" id="notes"></a></p><dl class="section user"><dt>Implementation Notes</dt><dd></dd></dl>
+<p>All functions (except <a class="el" href="array__ops_8sql__in.html#acb57ea4521dcb717f9e3148e0acccc74" title="This function normalizes an array as sum of squares to be 1. ">normalize()</a> and <a class="el" href="array__ops_8sql__in.html#acc295a568878940ffc3e2c9a75990efb" title="This function takes an array as the input and keep only elements that satisfy the operator on specifi...">array_filter()</a>) described in this module work with 2-D arrays.</p>
+<p>These functions support several numeric types:</p><ul>
+<li>SMALLINT</li>
+<li>INTEGER</li>
+<li>BIGINT</li>
+<li>REAL</li>
+<li>DOUBLE PRECISION (FLOAT8)</li>
+<li>NUMERIC (internally casted into FLOAT8, loss of precisions can happen)</li>
+</ul>
+<p>Several of the function require NO NULL VALUES, while others omit NULLs and return results. See details in description of individual functions.</p>
+<p><a class="anchor" id="list"></a></p><dl class="section user"><dt>Array Operations</dt><dd><table class="output">
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#a91c8d3715142927b3967f05a4fbf1575" title="Adds two arrays. It requires that all the values are NON-NULL. Return type is the same as the input t...">array_add()</a></th><td><p class="starttd">Adds two arrays. It requires that all the values are NON-NULL. Return type is the same as the input type.</p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#a26e8508a2bae10a6574cec697a270eea" title="Aggregate, element-wise sum of arrays. It requires that all the values are NON-NULL. Return type is the same as the input type. ">sum()</a></th><td><p class="starttd">Aggregate, sums vector element-wisely. It requires that all the values are NON-NULL. Return type is the same as the input type.</p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#a2875a161a01c7dcdea9a4997b074eefc" title="Subtracts two arrays. It requires that all the values are NON-NULL. Return type is the same as the in...">array_sub()</a></th><td><p class="starttd">Subtracts two arrays. It requires that all the values are NON-NULL. Return type is the same as the input type. </p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#a652d70c480d484c4a1a92ded384b0dd7" title="Element-wise product of two arrays. It requires that all the values are NON-NULL. Return type is the ...">array_mult()</a></th><td><p class="starttd">Element-wise product of two arrays. It requires that all the values are NON-NULL. Return type is the same as the input type. </p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#a6cc05e7052495f8b64692faf40219576" title="Element-wise division of two arrays. It requires that all the values are NON-NULL. Return type is the same as the input type. ">array_div()</a></th><td><p class="starttd">Element-wise division of two arrays. It requires that all the values are NON-NULL. Return type is the same as the input type. </p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#acde10964ed23b7c8da515fb84cb8d5e0" title="Dot-product of two arrays. It requires that all the values are NON-NULL. Return type is the same as t...">array_dot()</a></th><td><p class="starttd">Dot-product of two arrays. It requires that all the values are NON-NULL. Return type is the same as the input type. </p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#aedf6cb13eb4803bcc12dc4d95ea8ff4e" title="Checks whether one array contains the other. This function returns TRUE if each non-zero element in t...">array_contains()</a></th><td><p class="starttd">Checks whether one array contains the other. This function returns TRUE if each non-zero element in the right array equals to the element with the same index in the left array. </p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#ae891429cc50705c530f3e5ca15541849" title="This function finds the maximum value in the array. NULLs are ignored. Return type is the same as the...">array_max()</a></th><td><p class="starttd">This function finds the maximum value in the array. NULLs are ignored. Return type is the same as the input type. </p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#aa415256a9064aecc600dfb5e377fb7b1" title="This function finds the maximum value and corresponding index in the array. NULLs are ignored...">array_max_index()</a></th><td><p class="starttd">This function finds the maximum value and corresponding index in the array. NULLs are ignored. Return type is array in format [max, index], and its element type is the same as the input type. </p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#a6659bf9d9363eb179fab34f81f8ac59b" title="This function finds the minimum value in the array. NULLs are ignored. Return type is the same as the...">array_min()</a></th><td><p class="starttd">This function finds the minimum value in the array. NULLs are ignored. Return type is the same as the input type. </p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#a813a4d9ffc1c18b1b3e18f6ecdb2051f" title="This function finds the minimum value and corresponding index in the array. NULLs are ignored...">array_min_index()</a></th><td><p class="starttd">This function finds the minimum value and corresponding index in the array. NULLs are ignored. Return type is array in format [min, index], and its element type is the same as the input type. </p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#a4c98f20e6a737358806f63318daea5ec" title="This function finds the sum of the values in the array. NULLs are ignored. Return type is the same as...">array_sum()</a></th><td><p class="starttd">This function finds the sum of the values in the array. NULLs are ignored. Return type is the same as the input type. </p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#a418de59800833aa95f9b7cbd6b12901c" title="This function finds the sum of the values in the array. NULLs are ignored. Return type is always FLOA...">array_sum_big()</a></th><td><p class="starttd">This function finds the sum of the values in the array. NULLs are ignored. Return type is always FLOAT8 regardless of input. This function is meant to replace <a class="el" href="array__ops_8sql__in.html#a4c98f20e6a737358806f63318daea5ec" title="This function finds the sum of the values in the array. NULLs are ignored. Return type is the same as...">array_sum()</a> in cases when a sum may overflow the element type. </p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#a13c0b0c53e8b0dc4e08c21bb8152ee7d" title="This function finds the sum of abs of the values in the array. NULLs are ignored. Return type is the ...">array_abs_sum()</a></th><td><p class="starttd">This function finds the sum of abs of the values in the array. NULLs are ignored. Return type is the same as the input type. </p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#ac14e74c10b58f5518cd0e3e56067e5ba" title="This function takes an array as the input and finds absolute value of each element in the array...">array_abs()</a></th><td><p class="starttd">This function takes an array as the input and finds abs of each element in the array, returning the resulting array. It requires that all the values are NON-NULL. </p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#a407598f9eb70637798b02fd731bfca2c" title="This function finds the mean of the values in the array. NULLs are ignored. ">array_mean()</a></th><td><p class="starttd">This function finds the mean of the values in the array. NULLs are ignored. </p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#a3b6c2d173a611e6d6b184d825c2b336d" title="This function finds the standard deviation of the values in the array. NULLs are ignored. ">array_stddev()</a></th><td><p class="starttd">This function finds the standard deviation of the values in the array. NULLs are ignored. </p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#ab066e65a41db78b00b4532996b2a6efc" title="This function creates an array of set size (the argument value) of FLOAT8, initializing the values to...">array_of_float()</a></th><td><p class="starttd">This function creates an array of set size (the argument value) of FLOAT8, initializing the values to 0.0. </p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#ab7d8550e66d2e0bd54b8f0997d93880c" title="This function creates an array of set size (the argument value) of BIGINT, initializing the values to...">array_of_bigint()</a></th><td><p class="starttd">This function creates an array of set size (the argument value) of BIGINT, initializing the values to 0. </p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#a065a5323f3b742be47e39ad8b4c90fc2" title="This functions set every values in the array to some desired value (provided as the argument)...">array_fill()</a></th><td><p class="starttd">This functions set every value in the array to some desired value (provided as the argument). </p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#acc295a568878940ffc3e2c9a75990efb" title="This function takes an array as the input and keep only elements that satisfy the operator on specifi...">array_filter()</a></th><td><p class="starttd">This function takes an array as the input and keep only elements that satisfy the operator on specified scalar. It requires that the array is 1-D and all the values are NON-NULL. Return type is the same as the input type. By default, this function removes all zeros. </p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#ae6881cc5c86941b6ffca35d7f3cd5c12" title="This function takes an array as the input and executes element-wise multiplication by the scalar prov...">array_scalar_mult()</a></th><td><p class="starttd">This function takes an array as the input and executes element-wise multiplication by the scalar provided as the second argument, returning the resulting array. It requires that all the values are NON-NULL. Return type is the same as the input type. </p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#a0b6ffe59b12c3dee076c3059f9ab363f" title="This function takes an array as the input and executes element-wise addition by the scalar provided a...">array_scalar_add()</a></th><td><p class="starttd">This function takes an array as the input and executes element-wise addition of the scalar provided as the second argument, returning the resulting array. It requires that all the values are NON-NULL. Return type is the same as the input type. </p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#a83451ed0c3ca5a9c62751dba47e45df7" title="This function takes an array as the input and finds square root of each element in the array...">array_sqrt()</a></th><td><p class="starttd">This function takes an array as the input and finds square root of each element in the array, returning the resulting array. It requires that all the values are NON-NULL. </p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#a761e7ca753a5e1acf26896b37ed8b0bd" title="This function takes an array and a float8 as the input and finds power of each element in the array...">array_pow()</a></th><td><p class="starttd">This function takes an array and a float8 as the input and finds power of each element in the array, returning the resulting array. It requires that all the values are NON-NULL. </p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#aff60f4091bed6374683f047c8a70ef9a" title="This function takes an array as the input and finds square of each element in the array...">array_square()</a></th><td><p class="starttd">This function takes an array as the input and finds square of each element in the array, returning the resulting array. It requires that all the values are NON-NULL. </p>
+<p class="endtd"></p>
+</td></tr>
+<tr>
+<th><a class="el" href="array__ops_8sql__in.html#acb57ea4521dcb717f9e3148e0acccc74" title="This function normalizes an array as sum of squares to be 1. ">normalize()</a></th><td>This function normalizes an array as sum of squares to be 1. It requires that the array is 1-D and all the values are NON-NULL. </td></tr>
+</table>
+</dd></dl>
+<p><a class="anchor" id="examples"></a></p><dl class="section user"><dt>Examples</dt><dd></dd></dl>
+<ol type="1">
+<li>Create a database table with two integer array columns and add some data. <pre class="example">
+CREATE TABLE array_tbl ( id integer,
+ array1 integer[],
+ array2 integer[]
+ );
+INSERT INTO array_tbl VALUES
+ ( 1, '{1,2,3,4,5,6,7,8,9}', '{9,8,7,6,5,4,3,2,1}' ),
+ ( 2, '{1,1,0,1,1,2,3,99,8}','{0,0,0,-5,4,1,1,7,6}' );
+</pre></li>
+<li>Find the minimum, maximum, mean, and standard deviation of the <code>array1</code> column. <pre class="example">
+SELECT id, madlib.array_min(array1), madlib.array_max(array1),
+ madlib.array_min_index(array1), madlib.array_max_index(array1),
+ madlib.array_mean(array1), madlib.array_stddev(array1)
+FROM array_tbl;
+</pre> Result: <pre class="result">
+id | array_min | array_max | array_min_index | array_max_index | array_mean | array_stddev
+----+-----------+-----------+---------------+---------------+------------------+------------------
+ 1 | 1 | 9 | {1,1} | {9,9} | 5 | 2.73861278752583
+ 2 | 0 | 99 | {0,3} | {99,8} | 12.8888888888889 | 32.3784050118457(2 rows)
+</pre></li>
+<li>Perform array addition and subtraction. <pre class="example">
+SELECT id, madlib.array_add(array1,array2),
+ madlib.array_sub(array1,array2)
+FROM array_tbl;
+</pre> Result: <pre class="result">
+ id | array_add | array_sub
+ ---+------------------------------+-------------------------
+ 2 | {1,1,0,-4,5,3,4,106,14} | {1,1,0,6,-3,1,2,92,2}
+ 1 | {10,10,10,10,10,10,10,10,10} | {-8,-6,-4,-2,0,2,4,6,8}
+(2 rows)
+</pre></li>
+<li>Perform element-wise array multiplication and division. The row with <code>id=2</code> is excluded because the divisor array contains zero, which would cause a divide-by-zero error. <pre class="example">
+SELECT id, madlib.array_mult(array1,array2),
+ madlib.array_div(array1,array2)
+FROM array_tbl
+WHERE 0 != ALL(array2);</pre> Result: <pre class="result">
+ id | array_mult | array_div
+ ---+----------------------------+---------------------
+ 1 | {9,16,21,24,25,24,21,16,9} | {0,0,0,0,1,1,2,4,9}
+(1 row)
+</pre></li>
+<li>Calculate the dot product of the arrays. <pre class="example">
+SELECT id, madlib.array_dot(array1, array2)
+FROM array_tbl;
+</pre> Result: <pre class="result">
+ id | array_dot
+ ---+----------
+ 2 | 745
+ 1 | 165
+(2 rows)
+</pre></li>
+<li>Multiply an array by a scalar 3. <pre class="example">
+SELECT id, madlib.array_scalar_mult(array1,3)
+FROM array_tbl;
+</pre> Result: <pre class="result">
+ id | array_scalar_mult
+ ---+--------------------------
+ 1 | {3,6,9,12,15,18,21,24,27}
+ 2 | {3,3,0,3,3,6,9,297,24}
+(2 rows)
+</pre></li>
+<li>Construct a nine-element array with each element set to the value 1.3. <pre class="example">
+SELECT madlib.array_fill(madlib.array_of_float(9), 1.3::float);
+</pre> Result: <pre class="result">
+ array_fill
+ --------------------------------------
+ {1.3,1.3,1.3,1.3,1.3,1.3,1.3,1.3,1.3}
+(1 row)
+</pre></li>
+</ol>
+<p><a class="anchor" id="related"></a></p><dl class="section user"><dt>Related Topics</dt><dd></dd></dl>
+<p>File <a class="el" href="array__ops_8sql__in.html" title="implementation of array operations in SQL ">array_ops.sql_in</a> for list of functions and usage. </p>
+</div><!-- contents -->
+</div><!-- doc-content -->
+<!-- start footer part -->
+<div id="nav-path" class="navpath"><!-- id is needed for treeview function! -->
+ <ul>
+ <li class="footer">Generated on Tue Sep 20 2016 11:27:01 for MADlib by
+ <a href="http://www.doxygen.org/index.html">
+ <img class="footer" src="doxygen.png" alt="doxygen"/></a> 1.8.10 </li>
+ </ul>
+</div>
+</body>
+</html>
http://git-wip-us.apache.org/repos/asf/incubator-madlib-site/blob/bed9253d/docs/v1.9.1/group__grp__arraysmatrix.html
----------------------------------------------------------------------
diff --git a/docs/v1.9.1/group__grp__arraysmatrix.html b/docs/v1.9.1/group__grp__arraysmatrix.html
new file mode 100644
index 0000000..e7615a5
--- /dev/null
+++ b/docs/v1.9.1/group__grp__arraysmatrix.html
@@ -0,0 +1,170 @@
+<!-- HTML header for doxygen 1.8.4-->
+<!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 1.8.10"/>
+<meta name="keywords" content="madlib,postgres,greenplum,machine learning,data mining,deep learning,ensemble methods,data science,market basket analysis,affinity analysis,pca,lda,regression,elastic net,huber white,proportional hazards,k-means,latent dirichlet allocation,bayes,support vector machines,svm"/>
+<title>MADlib: Arrays and Matrices</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="navtreedata.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/searchdata.js"></script>
+<script type="text/javascript" src="search/search.js"></script>
+<script type="text/javascript">
+ $(document).ready(function() { init_search(); });
+</script>
+<!-- hack in the navigation tree -->
+<script type="text/javascript" src="eigen_navtree_hacks.js"></script>
+<link href="doxygen.css" rel="stylesheet" type="text/css" />
+<link href="madlib_extra.css" rel="stylesheet" type="text/css"/>
+<!-- google analytics -->
+<script>
+ (function(i,s,o,g,r,a,m){i['GoogleAnalyticsObject']=r;i[r]=i[r]||function(){
+ (i[r].q=i[r].q||[]).push(arguments)},i[r].l=1*new Date();a=s.createElement(o),
+ m=s.getElementsByTagName(o)[0];a.async=1;a.src=g;m.parentNode.insertBefore(a,m)
+ })(window,document,'script','//www.google-analytics.com/analytics.js','ga');
+ ga('create', 'UA-45382226-1', 'madlib.net');
+ ga('send', 'pageview');
+</script>
+</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 id="projectlogo"><a href="http://madlib.net"><img alt="Logo" src="madlib.png" height="50" style="padding-left:0.5em;" border="0"/ ></a></td>
+ <td style="padding-left: 0.5em;">
+ <div id="projectname">
+ <span id="projectnumber">1.9.1</span>
+ </div>
+ <div id="projectbrief">User Documentation for MADlib</div>
+ </td>
+ <td> <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>
+</td>
+ </tr>
+ </tbody>
+</table>
+</div>
+<!-- end header part -->
+<!-- Generated by Doxygen 1.8.10 -->
+<script type="text/javascript">
+var searchBox = new SearchBox("searchBox", "search",false,'Search');
+</script>
+</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('group__grp__arraysmatrix.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)">
+</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="summary">
+<a href="#groups">Modules</a> </div>
+ <div class="headertitle">
+<div class="title">Arrays and Matrices<div class="ingroups"><a class="el" href="group__grp__datatrans.html">Data Types and Transformations</a></div></div> </div>
+</div><!--header-->
+<div class="contents">
+<a name="details" id="details"></a><h2 class="groupheader">Detailed Description</h2>
+<p>These modules provide basic mathematical operations to be run on array and matrices.</p>
+<p>For a distributed system, a matrix cannot simply be represented as a 2D array of numbers in memory. <b>We provide two forms of distributed representation of a matrix</b>:</p>
+<ul>
+<li>Dense: The matrix is represented as a distributed collection of 1-D arrays. An example 3x10 matrix would be the below table: <pre>
+ row_id | row_vec
+--------+-------------------------
+ 1 | {9,6,5,8,5,6,6,3,10,8}
+ 2 | {8,2,2,6,6,10,2,1,9,9}
+ 3 | {3,9,9,9,8,6,3,9,5,6}
+</pre></li>
+<li>Sparse: The matrix is represented using the row and column indices for each non-zero entry of the matrix. Example: <pre>
+ row_id | col_id | value
+--------+--------+-------
+ 1 | 1 | 9
+ 1 | 5 | 6
+ 1 | 6 | 6
+ 2 | 1 | 8
+ 3 | 1 | 3
+ 3 | 2 | 9
+ 4 | 7 | 0
+(6 rows)
+</pre>   All matrix operations work with either form of representation.</li>
+</ul>
+<p>In many cases, a matrix function can be <b>decomposed to vector operations applied independently on each row of a matrix (or corresponding rows of two matrices)</b>. We have also provided access to these internal vector operations (<a class="el" href="group__grp__array.html">Array Operations</a>) for greater flexibility. Matrix operations like <em>matrix_add</em> use the corresponding vector operation (<em>array_add</em>) and also include additional validation and formating. Other functions like <em>matrix_mult</em> are complex and use a combination of such vector operations and other SQL operations.</p>
+<p><b>It's important to note</b> that these array functions are only available for the dense format representation of the matrix. In general, the scope of a single array function invocation is limited to only an array (1-dimensional or 2-dimensional) that fits in memory. When such function is executed on a table of arrays, the function is called multiple times - once for each array (or pair of arrays). On contrary, scope of a single matrix function invocation is the complete matrix stored as a distributed table. </p>
+<table class="memberdecls">
+<tr class="heading"><td colspan="2"><h2 class="groupheader"><a name="groups"></a>
+Modules</h2></td></tr>
+<tr class="memitem:group__grp__array"><td class="memItemLeft" align="right" valign="top"> </td><td class="memItemRight" valign="bottom"><a class="el" href="group__grp__array.html">Array Operations</a></td></tr>
+<tr class="memdesc:group__grp__array"><td class="mdescLeft"> </td><td class="mdescRight">Provides fast array operations supporting other MADlib modules. <br /></td></tr>
+<tr class="separator:"><td class="memSeparator" colspan="2"> </td></tr>
+<tr class="memitem:group__grp__matrix"><td class="memItemLeft" align="right" valign="top"> </td><td class="memItemRight" valign="bottom"><a class="el" href="group__grp__matrix.html">Matrix Operations</a></td></tr>
+<tr class="memdesc:group__grp__matrix"><td class="mdescLeft"> </td><td class="mdescRight">Provides fast matrix operations supporting other MADlib modules. <br /></td></tr>
+<tr class="separator:"><td class="memSeparator" colspan="2"> </td></tr>
+<tr class="memitem:group__grp__matrix__factorization"><td class="memItemLeft" align="right" valign="top"> </td><td class="memItemRight" valign="bottom"><a class="el" href="group__grp__matrix__factorization.html">Matrix Factorization</a></td></tr>
+<tr class="memdesc:group__grp__matrix__factorization"><td class="mdescLeft"> </td><td class="mdescRight">Matrix Factorization methods including Singular Value Decomposition and Low-rank Matrix Factorization. <br /></td></tr>
+<tr class="separator:"><td class="memSeparator" colspan="2"> </td></tr>
+<tr class="memitem:group__grp__linalg"><td class="memItemLeft" align="right" valign="top"> </td><td class="memItemRight" valign="bottom"><a class="el" href="group__grp__linalg.html">Norms and Distance functions</a></td></tr>
+<tr class="memdesc:group__grp__linalg"><td class="mdescLeft"> </td><td class="mdescRight">Provides utility functions for basic linear algebra operations. <br /></td></tr>
+<tr class="separator:"><td class="memSeparator" colspan="2"> </td></tr>
+<tr class="memitem:group__grp__svec"><td class="memItemLeft" align="right" valign="top"> </td><td class="memItemRight" valign="bottom"><a class="el" href="group__grp__svec.html">Sparse Vectors</a></td></tr>
+<tr class="memdesc:group__grp__svec"><td class="mdescLeft"> </td><td class="mdescRight">Implements a sparse vector data type that provides compressed storage of vectors that may have many duplicate elements. <br /></td></tr>
+<tr class="separator:"><td class="memSeparator" colspan="2"> </td></tr>
+</table>
+</div><!-- contents -->
+</div><!-- doc-content -->
+<!-- start footer part -->
+<div id="nav-path" class="navpath"><!-- id is needed for treeview function! -->
+ <ul>
+ <li class="footer">Generated on Tue Sep 20 2016 11:27:01 for MADlib by
+ <a href="http://www.doxygen.org/index.html">
+ <img class="footer" src="doxygen.png" alt="doxygen"/></a> 1.8.10 </li>
+ </ul>
+</div>
+</body>
+</html>
http://git-wip-us.apache.org/repos/asf/incubator-madlib-site/blob/bed9253d/docs/v1.9.1/group__grp__arraysmatrix.js
----------------------------------------------------------------------
diff --git a/docs/v1.9.1/group__grp__arraysmatrix.js b/docs/v1.9.1/group__grp__arraysmatrix.js
new file mode 100644
index 0000000..5e12094
--- /dev/null
+++ b/docs/v1.9.1/group__grp__arraysmatrix.js
@@ -0,0 +1,8 @@
+var group__grp__arraysmatrix =
+[
+ [ "Array Operations", "group__grp__array.html", null ],
+ [ "Matrix Operations", "group__grp__matrix.html", null ],
+ [ "Matrix Factorization", "group__grp__matrix__factorization.html", "group__grp__matrix__factorization" ],
+ [ "Norms and Distance functions", "group__grp__linalg.html", null ],
+ [ "Sparse Vectors", "group__grp__svec.html", null ]
+];
\ No newline at end of file
http://git-wip-us.apache.org/repos/asf/incubator-madlib-site/blob/bed9253d/docs/v1.9.1/group__grp__assoc__rules.html
----------------------------------------------------------------------
diff --git a/docs/v1.9.1/group__grp__assoc__rules.html b/docs/v1.9.1/group__grp__assoc__rules.html
new file mode 100644
index 0000000..9cc48c0
--- /dev/null
+++ b/docs/v1.9.1/group__grp__assoc__rules.html
@@ -0,0 +1,308 @@
+<!-- HTML header for doxygen 1.8.4-->
+<!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 1.8.10"/>
+<meta name="keywords" content="madlib,postgres,greenplum,machine learning,data mining,deep learning,ensemble methods,data science,market basket analysis,affinity analysis,pca,lda,regression,elastic net,huber white,proportional hazards,k-means,latent dirichlet allocation,bayes,support vector machines,svm"/>
+<title>MADlib: Apriori Algorithm</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="navtreedata.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/searchdata.js"></script>
+<script type="text/javascript" src="search/search.js"></script>
+<script type="text/javascript">
+ $(document).ready(function() { init_search(); });
+</script>
+<!-- hack in the navigation tree -->
+<script type="text/javascript" src="eigen_navtree_hacks.js"></script>
+<link href="doxygen.css" rel="stylesheet" type="text/css" />
+<link href="madlib_extra.css" rel="stylesheet" type="text/css"/>
+<!-- google analytics -->
+<script>
+ (function(i,s,o,g,r,a,m){i['GoogleAnalyticsObject']=r;i[r]=i[r]||function(){
+ (i[r].q=i[r].q||[]).push(arguments)},i[r].l=1*new Date();a=s.createElement(o),
+ m=s.getElementsByTagName(o)[0];a.async=1;a.src=g;m.parentNode.insertBefore(a,m)
+ })(window,document,'script','//www.google-analytics.com/analytics.js','ga');
+ ga('create', 'UA-45382226-1', 'madlib.net');
+ ga('send', 'pageview');
+</script>
+</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 id="projectlogo"><a href="http://madlib.net"><img alt="Logo" src="madlib.png" height="50" style="padding-left:0.5em;" border="0"/ ></a></td>
+ <td style="padding-left: 0.5em;">
+ <div id="projectname">
+ <span id="projectnumber">1.9.1</span>
+ </div>
+ <div id="projectbrief">User Documentation for MADlib</div>
+ </td>
+ <td> <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>
+</td>
+ </tr>
+ </tbody>
+</table>
+</div>
+<!-- end header part -->
+<!-- Generated by Doxygen 1.8.10 -->
+<script type="text/javascript">
+var searchBox = new SearchBox("searchBox", "search",false,'Search');
+</script>
+</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('group__grp__assoc__rules.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)">
+</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">Apriori Algorithm<div class="ingroups"><a class="el" href="group__grp__unsupervised.html">Unsupervised Learning</a> » <a class="el" href="group__grp__association__rules.html">Association Rules</a></div></div> </div>
+</div><!--header-->
+<div class="contents">
+<div class="toc"><b>Contents</b> </p><ul>
+<li>
+<a href="#rules">Rules</a> </li>
+<li>
+<a href="#algorithm">Apriori Algorithm</a> </li>
+<li>
+<a href="#syntax">Function Syntax</a> </li>
+<li>
+<a href="#examples">Examples</a> </li>
+<li>
+<a href="#notes">Notes</a> </li>
+<li>
+<a href="#related">Related Topics</a> </li>
+</ul>
+</div><p>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.</p>
+<p>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.</p>
+<p>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.</p>
+<pre>
+ tran_id | product
+ ---------+---------
+ 1 | 1
+ 1 | 2
+ 1 | 3
+ 1 | 4
+ 2 | 3
+ 2 | 4
+ 2 | 5
+ 3 | 1
+ 3 | 4
+ 3 | 6
+ ...
+</pre><p><a class="anchor" id="rules"></a></p><dl class="section user"><dt>Rules</dt><dd></dd></dl>
+<p>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.</p>
+<p>Given any association rule "If X, then Y", the association rules function will also calculate the following metrics:</p><ul>
+<li>Support: The ratio of transactions that contain X to all transactions, T <p class="formulaDsp">
+<img class="formulaDsp" alt="\[ S (X) = \frac{Total X}{Total transactions} \]" src="form_0.png"/>
+</p>
+</li>
+<li>Confidence: The ratio of transactions that contain <img class="formulaInl" alt="$ X,Y $" src="form_1.png"/> to transactions that contain <img class="formulaInl" alt="$ X $" src="form_2.png"/>. One could view this metric as the conditional probability of <img class="formulaInl" alt="$ Y $" src="form_3.png"/> , given <img class="formulaInl" alt="$ X $" src="form_2.png"/> . <img class="formulaInl" alt="$ P(Y|X) $" src="form_4.png"/> <p class="formulaDsp">
+<img class="formulaDsp" alt="\[ C (X \Rightarrow Y) = \frac{s(X \cap Y )}{s(X)} \]" src="form_5.png"/>
+</p>
+</li>
+<li>Lift: The ratio of observed support of <img class="formulaInl" alt="$ X,Y $" src="form_1.png"/> to the expected support of <img class="formulaInl" alt="$ X,Y $" src="form_1.png"/> , assuming <img class="formulaInl" alt="$ X $" src="form_2.png"/> and <img class="formulaInl" alt="$ Y $" src="form_3.png"/> are independent. <p class="formulaDsp">
+<img class="formulaDsp" alt="\[ L (X \Rightarrow Y) = \frac{s(X \cap Y )}{s(X) \cdot s(Y)} \]" src="form_6.png"/>
+</p>
+</li>
+<li><p class="startli">Conviction: The ratio of expected support of <img class="formulaInl" alt="$ X $" src="form_2.png"/> occurring without <img class="formulaInl" alt="$ Y $" src="form_3.png"/> assuming <img class="formulaInl" alt="$ X $" src="form_2.png"/> and <img class="formulaInl" alt="$ \neg Y $" src="form_7.png"/> are independent, to the observed support of <img class="formulaInl" alt="$ X $" src="form_2.png"/> occuring without <img class="formulaInl" alt="$ Y $" src="form_3.png"/>. If conviction is greater than 1, then this metric shows that incorrect predictions ( <img class="formulaInl" alt="$ X \Rightarrow Y $" src="form_8.png"/> ) 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.)</p>
+<p class="formulaDsp">
+<img class="formulaDsp" alt="\[ Conv (X \Rightarrow Y) = \frac{1 - S(Y)}{1 - C(X \Rightarrow Y)} \]" src="form_9.png"/>
+</p>
+</li>
+</ul>
+<p><a class="anchor" id="algorithm"></a></p><dl class="section user"><dt>Apriori Algorithm</dt><dd></dd></dl>
+<p>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 <img class="formulaInl" alt="$ n $" src="form_10.png"/> are generated from sets of order <img class="formulaInl" alt="$ n - 1 $" src="form_11.png"/>. Using the downward closure property, all sets must have frequent subsets. There are two steps in this algorithm; generating frequent itemsets, and using these itemsets to construct the association rules. A simplified version of the algorithm is as follows, and assumes a minimum level of support and confidence is provided:</p>
+<p><em>Initial</em> <em>step</em> </p><ol type="1">
+<li>Generate all itemsets of order 1</li>
+<li>Eliminate itemsets that have support is less than minimum support</li>
+</ol>
+<p><em>Main</em> <em>algorithm</em> </p><ol type="1">
+<li>For <img class="formulaInl" alt="$ n \ge 2 $" src="form_12.png"/>, generate itemsets of order <img class="formulaInl" alt="$ n $" src="form_10.png"/> by combining the itemsets of order <img class="formulaInl" alt="$ n - 1 $" src="form_11.png"/>. This is done by doing the union of two itemsets that have identical items except one.</li>
+<li>Eliminate itemsets that have (n-1) order subsets with insufficient support</li>
+<li>Eliminate itemsets with insufficient support</li>
+<li>Repeat until itemsets cannot be generated</li>
+</ol>
+<p><em>Association</em> <em>rule</em> <em>generation</em> </p>
+<p>Given a frequent itemset <img class="formulaInl" alt="$ A $" src="form_13.png"/> generated from the Apriori algorithm, and all subsets <img class="formulaInl" alt="$ B $" src="form_14.png"/> , we generate rules such that <img class="formulaInl" alt="$ B \Rightarrow (A - B) $" src="form_15.png"/> meets minimum confidence requirements.</p>
+<p><a class="anchor" id="syntax"></a></p><dl class="section user"><dt>Function Syntax</dt><dd>Association rules can be called with the following syntax. <pre class="syntax">
+assoc_rules( support,
+ confidence,
+ tid_col,
+ item_col,
+ input_table,
+ output_schema,
+ verbose
+ );</pre> This generates all association rules that satisfy the specified minimum <em>support</em> and <em>confidence</em>.</dd></dl>
+<p><b>Arguments</b> </p><dl class="arglist">
+<dt>support </dt>
+<dd><p class="startdd">The minimum level of support needed for each itemset to be included in result.</p>
+<p class="enddd"></p>
+</dd>
+<dt>confidence </dt>
+<dd><p class="startdd">The minimum level of confidence needed for each rule to be included in result.</p>
+<p class="enddd"></p>
+</dd>
+<dt>tid_col </dt>
+<dd><p class="startdd">The name of the column storing the transaction ids.</p>
+<p class="enddd"></p>
+</dd>
+<dt>item_col </dt>
+<dd><p class="startdd">The name of the column storing the products.</p>
+<p class="enddd"></p>
+</dd>
+<dt>input_table </dt>
+<dd><p class="startdd">The name of the table containing the input data.</p>
+<p>The input data is expected to be of the following form: </p><pre>{TABLE|VIEW} <em>input_table</em> (
+ <em>trans_id</em> INTEGER,
+ <em>product</em> TEXT
+)</pre><p>The algorithm maps the product names to consecutive integer ids starting at 1. If they are already structured this way, then the ids will not change. </p>
+<p class="enddd"></p>
+</dd>
+<dt>output_schema </dt>
+<dd><p class="startdd">The name of the schema where the final results will be stored. It is expected to be created before calling the function, or using <code>NULL</code> suggests the current schema will be used.</p>
+<p>The results containing the rules, support, confidence, lift, and conviction are stored in the table <code>assoc_rules</code> in the schema specified by <code>output_schema</code>.</p>
+<p>The table has the following columns. </p><table class="output">
+<tr>
+<th>ruleid </th><td>integer </td></tr>
+<tr>
+<th>pre </th><td>text </td></tr>
+<tr>
+<th>post </th><td>text </td></tr>
+<tr>
+<th>support </th><td>double </td></tr>
+<tr>
+<th>confidence </th><td>double </td></tr>
+<tr>
+<th>lift </th><td>double </td></tr>
+<tr>
+<th>conviction </th><td>double </td></tr>
+</table>
+<p>On Greenplum Database the table is distributed by the ruleid column.</p>
+<p>The <code>pre</code> and <code>post</code> columns are the itemsets of left and right hand sides of the association rule respectively. The <code>support</code>, <code>confidence</code>, <code>lift</code>, and <code>conviction</code> columns are calculated as mentioned in the about section. </p>
+<p class="enddd"></p>
+</dd>
+<dt>verbose </dt>
+<dd>BOOLEAN, default FALSE. Determines if the output contains comments. </dd>
+</dl>
+<p><a class="anchor" id="examples"></a></p><dl class="section user"><dt>Examples</dt><dd></dd></dl>
+<p>Let us take a look at some sample transactional data and generate association rules.</p>
+<ol type="1">
+<li>Create an input dataset. <pre class="example">
+DROP TABLE IF EXISTS test_data;
+CREATE TABLE test_data (
+ trans_id INT,
+ product TEXT
+);
+INSERT INTO test_data VALUES (1, 'beer');
+INSERT INTO test_data VALUES (1, 'diapers');
+INSERT INTO test_data VALUES (1, 'chips');
+INSERT INTO test_data VALUES (2, 'beer');
+INSERT INTO test_data VALUES (2, 'diapers');
+INSERT INTO test_data VALUES (3, 'beer');
+INSERT INTO test_data VALUES (3, 'diapers');
+INSERT INTO test_data VALUES (4, 'beer');
+INSERT INTO test_data VALUES (4, 'chips');
+INSERT INTO test_data VALUES (5, 'beer');
+INSERT INTO test_data VALUES (6, 'beer');
+INSERT INTO test_data VALUES (6, 'diapers');
+INSERT INTO test_data VALUES (6, 'chips');
+INSERT INTO test_data VALUES (7, 'beer');
+INSERT INTO test_data VALUES (7, 'diapers');
+</pre></li>
+<li>Let <img class="formulaInl" alt="$ min(support) = .25 $" src="form_16.png"/> and <img class="formulaInl" alt="$ min(confidence) = .5 $" src="form_17.png"/>, and the output schema be 'myschema'. For this example, we set verbose to TRUE so that we have some insight into the progress of the function. We can now generate association rules as follows: <pre class="example">
+SELECT * FROM madlib.assoc_rules( .25,
+ .5,
+ 'trans_id',
+ 'product',
+ 'test_data',
+ 'myschema',
+ TRUE
+ );
+</pre> Result: <pre class="result">
+ output_schema | output_table | total_rules | total_time
+---------------+--------------+-------------+-----------------
+ myschema | assoc_rules | 7 | 00:00:03.162094
+(1 row)
+</pre> The association rules are stored in the myschema.assoc_rules table: <pre class="example">
+SELECT * FROM myschema.assoc_rules
+ORDER BY support DESC;
+</pre> Result: <pre class="result">
+ ruleid | pre | post | support | confidence | lift | conviction
+--------+-----------------+----------------+-------------------+-------------------+-------------------+-------------------
+ 4 | {diapers} | {beer} | 0.714285714285714 | 1 | 1 | 0
+ 2 | {beer} | {diapers} | 0.714285714285714 | 0.714285714285714 | 1 | 1
+ 1 | {chips} | {beer} | 0.428571428571429 | 1 | 1 | 0
+ 5 | {chips} | {beer,diapers} | 0.285714285714286 | 0.666666666666667 | 0.933333333333333 | 0.857142857142857
+ 6 | {chips,beer} | {diapers} | 0.285714285714286 | 0.666666666666667 | 0.933333333333333 | 0.857142857142857
+ 7 | {chips,diapers} | {beer} | 0.285714285714286 | 1 | 1 | 0
+ 3 | {chips} | {diapers} | 0.285714285714286 | 0.666666666666667 | 0.933333333333333 | 0.857142857142857
+(7 rows)
+</pre></li>
+</ol>
+<p><a class="anchor" id="notes"></a></p><dl class="section user"><dt>Notes</dt><dd></dd></dl>
+<p>The association rules function always creates a table named <code>assoc_rules</code>. Make a copy of this table before running the function again if you would like to keep multiple association rule tables.</p>
+<p><a class="anchor" id="related"></a></p><dl class="section user"><dt>Related Topics</dt><dd></dd></dl>
+<p>File <a class="el" href="assoc__rules_8sql__in.html" title="The assoc_rules function computes association rules for a given set of data. The data is assumed to h...">assoc_rules.sql_in</a> documenting the SQL function. </p>
+</div><!-- contents -->
+</div><!-- doc-content -->
+<!-- start footer part -->
+<div id="nav-path" class="navpath"><!-- id is needed for treeview function! -->
+ <ul>
+ <li class="footer">Generated on Tue Sep 20 2016 11:27:01 for MADlib by
+ <a href="http://www.doxygen.org/index.html">
+ <img class="footer" src="doxygen.png" alt="doxygen"/></a> 1.8.10 </li>
+ </ul>
+</div>
+</body>
+</html>
http://git-wip-us.apache.org/repos/asf/incubator-madlib-site/blob/bed9253d/docs/v1.9.1/group__grp__association__rules.html
----------------------------------------------------------------------
diff --git a/docs/v1.9.1/group__grp__association__rules.html b/docs/v1.9.1/group__grp__association__rules.html
new file mode 100644
index 0000000..91fa398
--- /dev/null
+++ b/docs/v1.9.1/group__grp__association__rules.html
@@ -0,0 +1,134 @@
+<!-- HTML header for doxygen 1.8.4-->
+<!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 1.8.10"/>
+<meta name="keywords" content="madlib,postgres,greenplum,machine learning,data mining,deep learning,ensemble methods,data science,market basket analysis,affinity analysis,pca,lda,regression,elastic net,huber white,proportional hazards,k-means,latent dirichlet allocation,bayes,support vector machines,svm"/>
+<title>MADlib: Association Rules</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="navtreedata.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/searchdata.js"></script>
+<script type="text/javascript" src="search/search.js"></script>
+<script type="text/javascript">
+ $(document).ready(function() { init_search(); });
+</script>
+<!-- hack in the navigation tree -->
+<script type="text/javascript" src="eigen_navtree_hacks.js"></script>
+<link href="doxygen.css" rel="stylesheet" type="text/css" />
+<link href="madlib_extra.css" rel="stylesheet" type="text/css"/>
+<!-- google analytics -->
+<script>
+ (function(i,s,o,g,r,a,m){i['GoogleAnalyticsObject']=r;i[r]=i[r]||function(){
+ (i[r].q=i[r].q||[]).push(arguments)},i[r].l=1*new Date();a=s.createElement(o),
+ m=s.getElementsByTagName(o)[0];a.async=1;a.src=g;m.parentNode.insertBefore(a,m)
+ })(window,document,'script','//www.google-analytics.com/analytics.js','ga');
+ ga('create', 'UA-45382226-1', 'madlib.net');
+ ga('send', 'pageview');
+</script>
+</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 id="projectlogo"><a href="http://madlib.net"><img alt="Logo" src="madlib.png" height="50" style="padding-left:0.5em;" border="0"/ ></a></td>
+ <td style="padding-left: 0.5em;">
+ <div id="projectname">
+ <span id="projectnumber">1.9.1</span>
+ </div>
+ <div id="projectbrief">User Documentation for MADlib</div>
+ </td>
+ <td> <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>
+</td>
+ </tr>
+ </tbody>
+</table>
+</div>
+<!-- end header part -->
+<!-- Generated by Doxygen 1.8.10 -->
+<script type="text/javascript">
+var searchBox = new SearchBox("searchBox", "search",false,'Search');
+</script>
+</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('group__grp__association__rules.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)">
+</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="summary">
+<a href="#groups">Modules</a> </div>
+ <div class="headertitle">
+<div class="title">Association Rules<div class="ingroups"><a class="el" href="group__grp__unsupervised.html">Unsupervised Learning</a></div></div> </div>
+</div><!--header-->
+<div class="contents">
+<a name="details" id="details"></a><h2 class="groupheader">Detailed Description</h2>
+<p>A collection of methods used to uncover interesting patterns in transactional datasets. </p>
+<table class="memberdecls">
+<tr class="heading"><td colspan="2"><h2 class="groupheader"><a name="groups"></a>
+Modules</h2></td></tr>
+<tr class="memitem:group__grp__assoc__rules"><td class="memItemLeft" align="right" valign="top"> </td><td class="memItemRight" valign="bottom"><a class="el" href="group__grp__assoc__rules.html">Apriori Algorithm</a></td></tr>
+<tr class="memdesc:group__grp__assoc__rules"><td class="mdescLeft"> </td><td class="mdescRight">Computes association rules for a given set of data. <br /></td></tr>
+<tr class="separator:"><td class="memSeparator" colspan="2"> </td></tr>
+</table>
+</div><!-- contents -->
+</div><!-- doc-content -->
+<!-- start footer part -->
+<div id="nav-path" class="navpath"><!-- id is needed for treeview function! -->
+ <ul>
+ <li class="footer">Generated on Tue Sep 20 2016 11:27:01 for MADlib by
+ <a href="http://www.doxygen.org/index.html">
+ <img class="footer" src="doxygen.png" alt="doxygen"/></a> 1.8.10 </li>
+ </ul>
+</div>
+</body>
+</html>