| <!-- 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 1.8.13"/> |
| <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: Sparse Linear Systems</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); |
| </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> |
| <script type="text/x-mathjax-config"> |
| MathJax.Hub.Config({ |
| extensions: ["tex2jax.js", "TeX/AMSmath.js", "TeX/AMSsymbols.js"], |
| jax: ["input/TeX","output/HTML-CSS"], |
| }); |
| </script><script type="text/javascript" src="http://cdn.mathjax.org/mathjax/latest/MathJax.js"></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.apache.org'); |
| 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.apache.org"><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">1.19.0</span> |
| </div> |
| <div id="projectbrief">User Documentation for Apache 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 1.8.13 --> |
| <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__sparse__linear__solver.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">Sparse Linear Systems<div class="ingroups"><a class="el" href="group__grp__other__functions.html">Utilities</a> » <a class="el" href="group__grp__linear__solver.html">Linear Solvers</a></div></div> </div> |
| </div><!--header--> |
| <div class="contents"> |
| <div class="toc"><b>Contents</b> <ul> |
| <li class="level1"> |
| <a href="#sls_usage">Solution Function</a> </li> |
| <li class="level1"> |
| <a href="#sls_opt_params">Optimizer Parameters</a> </li> |
| <li class="level1"> |
| <a href="#sls_output">Output Tables</a> </li> |
| <li class="level1"> |
| <a href="#sls_examples">Examples</a> </li> |
| <li> |
| <a href="related">Related Topics</a> </li> |
| </ul> |
| </div><p>The sparse linear systems module implements solution methods for systems of consistent linear equations. Systems of linear equations take the form: </p><p class="formulaDsp"> |
| \[ Ax = b \] |
| </p> |
| <p>where \(x \in \mathbb{R}^{n}\), \(A \in \mathbb{R}^{m \times n} \) and \(b \in \mathbb{R}^{m}\). This module accepts sparse matrix input formats for \(A\) and \(b\). We assume that there are no rows of \(A\) where all elements are zero.</p> |
| <dl class="section note"><dt>Note</dt><dd>Algorithms with fail if there is an row of the input matrix containing all zeros.</dd></dl> |
| <p>The algorithms implemented in this module can handle large sparse square linear systems. Currently, the algorithms implemented in this module solve the linear system using direct or iterative methods.</p> |
| <p><a class="anchor" id="sls_usage"></a></p><dl class="section user"><dt>Sparse Linear Systems Solution Function</dt><dd></dd></dl> |
| <pre class="syntax"> |
| linear_solver_sparse( tbl_source_lhs, |
| tbl_source_rhs, |
| tbl_result, |
| lhs_row_id, |
| lhs_col_id, |
| lhs_value, |
| rhs_row_id, |
| rhs_value, |
| grouping_cols := NULL, |
| optimizer := 'direct', |
| optimizer_params := |
| 'algorithm = llt' |
| ) |
| </pre><p> <b>Arguments</b> </p><dl class="arglist"> |
| <dt>tbl_source_lhs </dt> |
| <dd><p class="startdd">The name of the table containing the left hand side matrix. For the LHS matrix, the input data is expected to be of the following form: </p><pre> |
| {TABLE|VIEW} <em>sourceName</em> ( |
| ... |
| <em>row_id</em> FLOAT8, |
| <em>col_id</em> FLOAT8, |
| <em>value</em> FLOAT8, |
| ... |
| )</pre><p> Each row represents a single equation. The <em>rhs</em> columns refer to the right hand side of the equations and the <em>lhs</em> columns refer to the multipliers on the variables on the left hand side of the same equations. </p> |
| <p class="enddd"></p> |
| </dd> |
| <dt>tbl_source_rhs </dt> |
| <dd><p class="startdd">TEXT. The name of the table containing the right hand side vector. For the RHS matrix, the input data is expected to be of the following form: </p><pre class="fragment">{TABLE|VIEW} <em>sourceName</em> ( |
| ... |
| <em>row_id</em> FLOAT8, |
| <em>value</em> FLOAT8 |
| ... |
| )</pre><p> Each row represents a single equation. The <em>rhs</em> columns refer to the right hand side of the equations while the <em>lhs</em> columns refers to the multipliers on the variables on the left hand side of the same equations. </p> |
| <p class="enddd"></p> |
| </dd> |
| <dt>tbl_result </dt> |
| <dd><p class="startdd">TEXT. The name of the table where the output is saved. Output is stored in the tabled named by the <em>tbl_result</em> argument. The table contains the following columns. The output contains the following columns: </p><table class="output"> |
| <tr> |
| <th>solution </th><td>FLOAT8[]. The solution is an array with the variables in the same order as that provided as input in the 'left_hand_side' column name of the 'source_table' </td></tr> |
| <tr> |
| <th>residual_norm </th><td>FLOAT8. Scaled residual norm, defined as \( \frac{|Ax - b|}{|b|} \). This value is an indication of the accuracy of the solution. </td></tr> |
| <tr> |
| <th>iters </th><td>INTEGER. Number of iterations required by the algorithm (only applicable for iterative algorithms) . The output is NULL for 'direct' methods. </td></tr> |
| </table> |
| <p class="enddd"></p> |
| </dd> |
| <dt>lhs_row_id </dt> |
| <dd>TEXT. The name of the column storing the 'row id' of the equations. <dl class="section note"><dt>Note</dt><dd>For a system with N equations, the row_id's must be a continuous range of integers from \( 0 \ldots n-1 \).</dd></dl> |
| </dd> |
| <dt>lhs_col_id </dt> |
| <dd><p class="startdd">TEXT. The name of the column (in tbl_source_lhs) storing the 'col id' of the equations.</p> |
| <p class="enddd"></p> |
| </dd> |
| <dt>lhs_value </dt> |
| <dd><p class="startdd">TEXT. The name of the column (in tbl_source_lhs) storing the 'value' of the equations.</p> |
| <p class="enddd"></p> |
| </dd> |
| <dt>rhs_row_id </dt> |
| <dd><p class="startdd">TEXT. The name of the column (in tbl_source_rhs) storing the 'col id' of the equations.</p> |
| <p class="enddd"></p> |
| </dd> |
| <dt>rhs_value </dt> |
| <dd><p class="startdd">TEXT. The name of the column (in tbl_source_rhs) storing the 'value' of the equations.</p> |
| <p class="enddd"></p> |
| </dd> |
| <dt>num_vars </dt> |
| <dd><p class="startdd">INTEGER. The number of variables in the linear system equations.</p> |
| <p class="enddd"></p> |
| </dd> |
| <dt>grouping_col (optional) </dt> |
| <dd>TEXT, default: NULL. Group by column names. <dl class="section note"><dt>Note</dt><dd>The grouping feature is currently not implemented and this parameter is only a placeholder.</dd></dl> |
| </dd> |
| <dt>optimizer (optional) </dt> |
| <dd><p class="startdd">TEXT, default: 'direct'. Type of optimizer.</p> |
| <p class="enddd"></p> |
| </dd> |
| <dt>optimizer_params (optional) </dt> |
| <dd>TEXT, default: NULL. Optimizer specific parameters. </dd> |
| </dl> |
| <p><a class="anchor" id="sls_opt_params"></a></p><dl class="section user"><dt>Optimizer Parameters</dt><dd></dd></dl> |
| <p>For each optimizer, there are specific parameters that can be tuned for better performance.</p> |
| <dl class="arglist"> |
| <dt>algorithm (default: ldlt) </dt> |
| <dd><p class="startdd"></p> |
| <p>There are several algorithms that can be classified as 'direct' methods of solving linear systems. Madlib functions provide various algorithmic options available for users.</p> |
| <p>The following table provides a guideline on the choice of algorithm based on conditions on the A matrix, speed of the algorithms and numerical stability.</p> |
| <pre class="fragment"> Algorithm | Conditions on A | Speed | Memory |
| ---------------------------------------------------------- |
| llt | Sym. Pos Def | ++ | --- |
| ldlt | Sym. Pos Def | ++ | --- |
| |
| For speed '++' is faster than '+', which is faster than '-'. |
| For accuracy '+++' is better than '++'. |
| For memory, '-' uses less memory than '--'. |
| |
| Note: ldlt is often preferred over llt |
| </pre><p>There are several algorithms that can be classified as 'iterative' methods of solving linear systems. Madlib functions provide various algorithmic options available for users.</p> |
| <p>The following table provides a guideline on the choice of algorithm based on conditions on the A matrix, speed of the algorithms and numerical stability.</p> |
| <pre class="fragment"> Algorithm | Conditions on A | Speed | Memory | Convergence |
| ---------------------------------------------------------------------- |
| cg-mem | Sym. Pos Def | +++ | - | ++ |
| bicgstab-mem | Square | ++ | - | + |
| precond-cg-mem | Sym. Pos Def | ++ | - | +++ |
| precond-bicgstab-mem | Square | + | - | ++ |
| |
| For memory, '-' uses less memory than '--'. |
| For speed, '++' is faster than '+'. |
| </pre><p>Algorithm Details: </p><table class="output"> |
| <tr> |
| <th>cg-mem</th><td>In memory conjugate gradient with diagonal preconditioners. </td></tr> |
| <tr> |
| <th>bicgstab-mem</th><td>Bi-conjugate gradient (equivalent to performing CG on the least squares formulation of Ax=b) with incomplete LU preconditioners. </td></tr> |
| <tr> |
| <th>precond-cg-mem</th><td>In memory conjugate gradient with diagonal preconditioners. </td></tr> |
| <tr> |
| <th>bicgstab-mem</th><td>Bi-conjugate gradient (equivalent to performing CG on the least squares formulation of Ax=b) with incomplete LU preconditioners. </td></tr> |
| </table> |
| <p class="enddd"></p> |
| </dd> |
| <dt>toler (default: 1e-5) </dt> |
| <dd><p class="startdd">Termination tolerance (applicable only for iterative methods) which determines the stopping criterion (with respect to residual norm) for iterative methods. </p> |
| <p class="enddd"></p> |
| </dd> |
| </dl> |
| <p><a class="anchor" id="sls_examples"></a></p><dl class="section user"><dt>Examples</dt><dd></dd></dl> |
| <ol type="1"> |
| <li>View online help for the sparse linear systems solver function. <pre class="example"> |
| SELECT madlib.linear_solver_sparse(); |
| </pre></li> |
| <li>Create the sample data set. <pre class="example"> |
| DROP TABLE IF EXISTS sparse_linear_systems_lhs; |
| CREATE TABLE sparse_linear_systems_lhs ( |
| rid INTEGER NOT NULL, |
| cid INTEGER, |
| val DOUBLE PRECISION |
| ); |
| DROP TABLE IF EXISTS sparse_linear_systems_rhs; |
| CREATE TABLE sparse_linear_systems_rhs ( |
| rid INTEGER NOT NULL, |
| val DOUBLE PRECISION |
| ); |
| INSERT INTO sparse_linear_systems_lhs(rid, cid, val) VALUES |
| (0, 0, 1), |
| (1, 1, 1), |
| (2, 2, 1), |
| (3, 3, 1); |
| INSERT INTO sparse_linear_systems_rhs(rid, val) VALUES |
| (0, 10), |
| (1, 20), |
| (2, 30); |
| </pre></li> |
| <li>Solve the linear systems with default parameters. <pre class="example"> |
| SELECT madlib.linear_solver_sparse( 'sparse_linear_systems_lhs', |
| 'sparse_linear_systems_rhs', |
| 'output_table', |
| 'rid', |
| 'cid', |
| 'val', |
| 'rid', |
| 'val', |
| 4 |
| ); |
| </pre></li> |
| <li>View the contents of the output table. <pre class="example"> |
| \x on |
| SELECT * FROM output_table; |
| </pre> Result: <pre class="result"> |
| --------------------+------------------------------------- |
| solution | {10,20,30,0} |
| residual_norm | 0 |
| iters | NULL |
| </pre></li> |
| <li>Choose a different algorithm than the default algorithm. <pre class="example"> |
| DROP TABLE IF EXISTS output_table; |
| SELECT madlib.linear_solver_sparse( 'sparse_linear_systems_lhs', |
| 'sparse_linear_systems_rhs', |
| 'output_table', |
| 'rid', |
| 'cid', |
| 'val', |
| 'rid', |
| 'val', |
| 4, |
| NULL, |
| 'direct', |
| 'algorithm=llt' |
| ); |
| </pre></li> |
| <li>Choose a different algorithm than the default algorithm. <pre class="example"> |
| DROP TABLE IF EXISTS output_table; |
| SELECT madlib.linear_solver_sparse( |
| 'sparse_linear_systems_lhs', |
| 'sparse_linear_systems_rhs', |
| 'output_table', |
| 'rid', |
| 'cid', |
| 'val', |
| 'rid', |
| 'val', |
| 4, |
| NULL, |
| 'iterative', |
| 'algorithm=cg-mem, toler=1e-5' |
| ); |
| </pre></li> |
| </ol> |
| <p><a class="anchor" id="related"></a></p><dl class="section user"><dt>Related Topics</dt><dd>File sparse_linear_sytems.sql_in documenting the SQL functions.</dd></dl> |
| </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 Wed Dec 15 2021 20:27:21 for MADlib by |
| <a href="http://www.doxygen.org/index.html"> |
| <img class="footer" src="doxygen.png" alt="doxygen"/></a> 1.8.13 </li> |
| </ul> |
| </div> |
| </body> |
| </html> |