<?xml version="1.0" encoding="UTF-8" standalone="no"?>
<!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/html; charset=UTF-8" /><title>12.9. GIN and GiST Index Types</title><link rel="stylesheet" type="text/css" href="stylesheet.css" /><link rev="made" href="pgsql-docs@lists.postgresql.org" /><meta name="generator" content="DocBook XSL Stylesheets Vsnapshot" /><link rel="prev" href="textsearch-debugging.html" title="12.8. Testing and Debugging Text Search" /><link rel="next" href="textsearch-psql.html" title="12.10. psql Support" /></head><body><div xmlns="http://www.w3.org/TR/xhtml1/transitional" class="navheader"><table width="100%" summary="Navigation header"><tr><th colspan="5" align="center">12.9. GIN and GiST Index Types</th></tr><tr><td width="10%" align="left"><a accesskey="p" href="textsearch-debugging.html" title="12.8. Testing and Debugging Text Search">Prev</a> </td><td width="10%" align="left"><a accesskey="u" href="textsearch.html" title="Chapter 12. Full Text Search">Up</a></td><th width="60%" align="center">Chapter 12. Full Text Search</th><td width="10%" align="right"><a accesskey="h" href="index.html" title="PostgreSQL 11.5 Documentation">Home</a></td><td width="10%" align="right"> <a accesskey="n" href="textsearch-psql.html" title="12.10. psql Support">Next</a></td></tr></table><hr></hr></div><div class="sect1" id="TEXTSEARCH-INDEXES"><div class="titlepage"><div><div><h2 class="title" style="clear: both">12.9. GIN and GiST Index Types</h2></div></div></div><a id="id-1.5.11.12.2" class="indexterm"></a><p>
There are two kinds of indexes that can be used to speed up full text
searches.
Note that indexes are not mandatory for full text searching, but in
cases where a column is searched on a regular basis, an index is
usually desirable.
</p><div class="variablelist"><dl class="variablelist"><dt><span class="term">
<a id="id-1.5.11.12.3.1.1.1.1" class="indexterm"></a>
<code class="literal">CREATE INDEX <em class="replaceable"><code>name</code></em> ON <em class="replaceable"><code>table</code></em> USING GIN (<em class="replaceable"><code>column</code></em>);</code>
</span></dt><dd><p>
Creates a GIN (Generalized Inverted Index)-based index.
The <em class="replaceable"><code>column</code></em> must be of <code class="type">tsvector</code> type.
</p></dd><dt><span class="term">
<a id="id-1.5.11.12.3.1.2.1.1" class="indexterm"></a>
<code class="literal">CREATE INDEX <em class="replaceable"><code>name</code></em> ON <em class="replaceable"><code>table</code></em> USING GIST (<em class="replaceable"><code>column</code></em>);</code>
</span></dt><dd><p>
Creates a GiST (Generalized Search Tree)-based index.
The <em class="replaceable"><code>column</code></em> can be of <code class="type">tsvector</code> or
<code class="type">tsquery</code> type.
</p></dd></dl></div><p>
</p><p>
GIN indexes are the preferred text search index type. As inverted
indexes, they contain an index entry for each word (lexeme), with a
compressed list of matching locations. Multi-word searches can find
the first match, then use the index to remove rows that are lacking
additional words. GIN indexes store only the words (lexemes) of
<code class="type">tsvector</code> values, and not their weight labels. Thus a table
row recheck is needed when using a query that involves weights.
</p><p>
A GiST index is <em class="firstterm">lossy</em>, meaning that the index
might produce false matches, and it is necessary
to check the actual table row to eliminate such false matches.
(<span class="productname">PostgreSQL</span> does this automatically when needed.)
GiST indexes are lossy because each document is represented in the
index by a fixed-length signature. The signature is generated by hashing
each word into a single bit in an n-bit string, with all these bits OR-ed
together to produce an n-bit document signature. When two words hash to
the same bit position there will be a false match. If all words in
the query have matches (real or false) then the table row must be
retrieved to see if the match is correct.
</p><p>
Lossiness causes performance degradation due to unnecessary fetches of table
records that turn out to be false matches. Since random access to table
records is slow, this limits the usefulness of GiST indexes. The
likelihood of false matches depends on several factors, in particular the
number of unique words, so using dictionaries to reduce this number is
recommended.
</p><p>
Note that <acronym class="acronym">GIN</acronym> index build time can often be improved
by increasing <a class="xref" href="runtime-config-resource.html#GUC-MAINTENANCE-WORK-MEM">maintenance_work_mem</a>, while
<acronym class="acronym">GiST</acronym> index build time is not sensitive to that
parameter.
</p><p>
Partitioning of big collections and the proper use of GIN and GiST indexes
allows the implementation of very fast searches with online update.
Partitioning can be done at the database level using table inheritance,
or by distributing documents over
servers and collecting external search results, e.g. via <a class="link" href="ddl-foreign-data.html" title="5.11. Foreign Data">Foreign Data</a> access.
The latter is possible because ranking functions use
only local information.
</p></div><div class="navfooter"><hr /><table width="100%" summary="Navigation footer"><tr><td width="40%" align="left"><a accesskey="p" href="textsearch-debugging.html">Prev</a> </td><td width="20%" align="center"><a accesskey="u" href="textsearch.html">Up</a></td><td width="40%" align="right"> <a accesskey="n" href="textsearch-psql.html">Next</a></td></tr><tr><td width="40%" align="left" valign="top">12.8. Testing and Debugging Text Search </td><td width="20%" align="center"><a accesskey="h" href="index.html">Home</a></td><td width="40%" align="right" valign="top"> 12.10. <span class="application">psql</span> Support</td></tr></table></div></body></html>
