2 $PostgreSQL: pgsql/doc/src/sgml/ref/create_index.sgml,v 1.52 2005/11/07 17:36:44 tgl Exp $
3 PostgreSQL documentation
6 <refentry id="SQL-CREATEINDEX">
8 <refentrytitle id="sql-createindex-title">CREATE INDEX</refentrytitle>
9 <refmiscinfo>SQL - Language Statements</refmiscinfo>
13 <refname>CREATE INDEX</refname>
14 <refpurpose>define a new index</refpurpose>
17 <indexterm zone="sql-createindex">
18 <primary>CREATE INDEX</primary>
23 CREATE [ UNIQUE ] INDEX <replaceable class="parameter">name</replaceable> ON <replaceable class="parameter">table</replaceable> [ USING <replaceable class="parameter">method</replaceable> ]
24 ( { <replaceable class="parameter">column</replaceable> | ( <replaceable class="parameter">expression</replaceable> ) } [ <replaceable class="parameter">opclass</replaceable> ] [, ...] )
25 [ TABLESPACE <replaceable class="parameter">tablespace</replaceable> ]
26 [ WHERE <replaceable class="parameter">predicate</replaceable> ]
31 <title>Description</title>
34 <command>CREATE INDEX</command> constructs an index <replaceable
35 class="parameter">index_name</replaceable> on the specified table.
36 Indexes are primarily used to enhance database performance (though
37 inappropriate use can result in slower performance).
41 The key field(s) for the index are specified as column names,
42 or alternatively as expressions written in parentheses.
43 Multiple fields can be specified if the index method supports
48 An index field can be an expression computed from the values of
49 one or more columns of the table row. This feature can be used
50 to obtain fast access to data based on some transformation of
51 the basic data. For example, an index computed on
52 <literal>upper(col)</> would allow the clause
53 <literal>WHERE upper(col) = 'JIM'</> to use an index.
57 <productname>PostgreSQL</productname> provides the index methods
58 B-tree, hash, and GiST. Users can also define their own index
59 methods, but that is fairly complicated.
63 When the <literal>WHERE</literal> clause is present, a
64 <firstterm>partial index</firstterm> is created.
65 A partial index is an index that contains entries for only a portion of
66 a table, usually a portion that is more useful for indexing than the
67 rest of the table. For example, if you have a table that contains both
68 billed and unbilled orders where the unbilled orders take up a small
69 fraction of the total table and yet that is an often used section, you
70 can improve performance by creating an index on just that portion.
71 Another possible application is to use <literal>WHERE</literal> with
72 <literal>UNIQUE</literal> to enforce uniqueness over a subset of a
73 table. See <xref linkend="indexes-partial"> for more discussion.
77 The expression used in the <literal>WHERE</literal> clause may refer
78 only to columns of the underlying table, but it can use all columns,
79 not just the ones being indexed. Presently, subqueries and
80 aggregate expressions are also forbidden in <literal>WHERE</literal>.
81 The same restrictions apply to index fields that are expressions.
85 All functions and operators used in an index definition must be
86 <quote>immutable</>, that is, their results must depend only on
87 their arguments and never on any outside influence (such as
88 the contents of another table or the current time). This restriction
89 ensures that the behavior of the index is well-defined. To use a
90 user-defined function in an index expression or <literal>WHERE</literal>
91 clause, remember to mark the function immutable when you create it.
96 <title>Parameters</title>
100 <term><literal>UNIQUE</literal></term>
103 Causes the system to check for
104 duplicate values in the table when the index is created (if data
105 already exist) and each time data is added. Attempts to
106 insert or update data which would result in duplicate entries
107 will generate an error.
113 <term><replaceable class="parameter">name</replaceable></term>
116 The name of the index to be created. No schema name can be included
117 here; the index is always created in the same schema as its parent
124 <term><replaceable class="parameter">table</replaceable></term>
127 The name (possibly schema-qualified) of the table to be indexed.
133 <term><replaceable class="parameter">method</replaceable></term>
136 The name of the index method to be used. Choices are
137 <literal>btree</literal>, <literal>hash</literal>,
138 and <literal>gist</literal>. The
139 default method is <literal>btree</literal>.
145 <term><replaceable class="parameter">column</replaceable></term>
148 The name of a column of the table.
154 <term><replaceable class="parameter">expression</replaceable></term>
157 An expression based on one or more columns of the table. The
158 expression usually must be written with surrounding parentheses,
159 as shown in the syntax. However, the parentheses may be omitted
160 if the expression has the form of a function call.
166 <term><replaceable class="parameter">opclass</replaceable></term>
169 The name of an operator class. See below for details.
175 <term><replaceable class="parameter">tablespace</replaceable></term>
178 The tablespace in which to create the index. If not specified,
179 <xref linkend="guc-default-tablespace"> is used, or the database's
180 default tablespace if <varname>default_tablespace</> is an empty
187 <term><replaceable class="parameter">predicate</replaceable></term>
190 The constraint expression for a partial index.
202 See <xref linkend="indexes"> for information about when indexes can
203 be used, when they are not used, and in which particular situations
208 Currently, only the B-tree and GiST index methods support
209 multicolumn indexes. Up to 32 fields may be specified by default.
210 (This limit can be altered when building
211 <productname>PostgreSQL</productname>.) Only B-tree currently
212 supports unique indexes.
216 An <firstterm>operator class</firstterm> can be specified for each
217 column of an index. The operator class identifies the operators to be
218 used by the index for that column. For example, a B-tree index on
219 four-byte integers would use the <literal>int4_ops</literal> class;
220 this operator class includes comparison functions for four-byte
221 integers. In practice the default operator class for the column's data
222 type is usually sufficient. The main point of having operator classes
223 is that for some data types, there could be more than one meaningful
224 ordering. For example, we might want to sort a complex-number data
225 type either by absolute value or by real part. We could do this by
226 defining two operator classes for the data type and then selecting
227 the proper class when making an index. More information about
228 operator classes is in <xref linkend="indexes-opclass"> and in <xref
233 Use <xref linkend="sql-dropindex" endterm="sql-dropindex-title">
238 Indexes are not used for <literal>IS NULL</> clauses by default.
239 The best way to use indexes in such cases is to create a partial index
240 using an <literal>IS NULL</> predicate.
244 Prior releases of <productname>PostgreSQL</productname> also had an
245 R-tree index method. This method has been removed because
246 it had no significant advantages over the GiST method.
247 If <literal>USING rtree</> is specified, <command>CREATE INDEX</>
248 will interpret it as <literal>USING gist</>, to simplify conversion
249 of old databases to GiST.
254 <title>Examples</title>
257 To create a B-tree index on the column <literal>title</literal> in
258 the table <literal>films</literal>:
260 CREATE UNIQUE INDEX title_idx ON films (title);
265 To create an index on the column <literal>code</> in the table
266 <literal>films</> and have the index reside in the tablespace
267 <literal>indexspace</>:
269 CREATE INDEX code_idx ON films(code) TABLESPACE indexspace;
275 Is this example correct?
278 To create a GiST index on a point attribute so that we
279 can efficiently use box operators on the result of the
283 CREATE INDEX pointloc
284 ON points USING GIST (point2box(location) box_ops);
286 WHERE point2box(points.pointloc) = boxes.box;
293 <title>Compatibility</title>
296 <command>CREATE INDEX</command> is a
297 <productname>PostgreSQL</productname> language extension. There
298 are no provisions for indexes in the SQL standard.
303 <title>See Also</title>
305 <simplelist type="inline">
306 <member><xref linkend="sql-alterindex" endterm="sql-alterindex-title"></member>
307 <member><xref linkend="sql-dropindex" endterm="sql-dropindex-title"></member>
312 <!-- Keep this comment at the end of the file
317 sgml-minimize-attributes:nil
318 sgml-always-quote-attributes:t
321 sgml-parent-document:nil
322 sgml-default-dtd-file:"../reference.ced"
323 sgml-exposed-tags:nil
324 sgml-local-catalogs:"/usr/lib/sgml/catalog"
325 sgml-local-ecat-files:nil