Baoding-FenShuaJiang
/
Green_GoodERP18_FSJ_Standard
Watch 1
Star 0
Wiki 捐赠
gooderp18绿色标准版
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
Tree: 0f75e31513
Branches Tags
${ item.name }
Create branch ${ searchTerm }
from '0f75e31513'
${ noResults }
Green_GoodERP18_FSJ_Standard/PostgreSQL/doc/postgresql/html/functions-aggregate.html

563 lines
39KB
Raw Blame History 

  1. <?xml version="1.0" encoding="UTF-8" standalone="no"?>

  2. <!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>9.20. Aggregate Functions</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 V1.79.1" /><link rel="prev" href="functions-range.html" title="9.19. Range Functions and Operators" /><link rel="next" href="functions-window.html" title="9.21. Window Functions" /></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">9.20. Aggregate Functions</th></tr><tr><td width="10%" align="left"><a accesskey="p" href="functions-range.html" title="9.19. Range Functions and Operators">Prev</a> </td><td width="10%" align="left"><a accesskey="u" href="functions.html" title="Chapter 9. Functions and Operators">Up</a></td><th width="60%" align="center">Chapter 9. Functions and Operators</th><td width="10%" align="right"><a accesskey="h" href="index.html" title="PostgreSQL 12.4 Documentation">Home</a></td><td width="10%" align="right"> <a accesskey="n" href="functions-window.html" title="9.21. Window Functions">Next</a></td></tr></table><hr></hr></div><div class="sect1" id="FUNCTIONS-AGGREGATE"><div class="titlepage"><div><div><h2 class="title" style="clear: both">9.20. Aggregate Functions</h2></div></div></div><a id="id-1.5.8.25.2" class="indexterm"></a><p>

  3.    <em class="firstterm">Aggregate functions</em> compute a single result

  4.    from a set of input values.  The built-in general-purpose aggregate

  5.    functions are listed in <a class="xref" href="functions-aggregate.html#FUNCTIONS-AGGREGATE-TABLE" title="Table 9.55. General-Purpose Aggregate Functions">Table 9.55</a>

  6.    and statistical aggregates in <a class="xref" href="functions-aggregate.html#FUNCTIONS-AGGREGATE-STATISTICS-TABLE" title="Table 9.56. Aggregate Functions for Statistics">Table 9.56</a>.

  7.    The built-in within-group ordered-set aggregate functions

  8.    are listed in <a class="xref" href="functions-aggregate.html#FUNCTIONS-ORDEREDSET-TABLE" title="Table 9.57. Ordered-Set Aggregate Functions">Table 9.57</a>

  9.    while the built-in within-group hypothetical-set ones are in <a class="xref" href="functions-aggregate.html#FUNCTIONS-HYPOTHETICAL-TABLE" title="Table 9.58. Hypothetical-Set Aggregate Functions">Table 9.58</a>.  Grouping operations,

  10.    which are closely related to aggregate functions, are listed in

  11.    <a class="xref" href="functions-aggregate.html#FUNCTIONS-GROUPING-TABLE" title="Table 9.59. Grouping Operations">Table 9.59</a>.

  12.    The special syntax considerations for aggregate

  13.    functions are explained in <a class="xref" href="sql-expressions.html#SYNTAX-AGGREGATES" title="4.2.7. Aggregate Expressions">Section 4.2.7</a>.

  14.    Consult <a class="xref" href="tutorial-agg.html" title="2.7. Aggregate Functions">Section 2.7</a> for additional introductory

  15.    information.

  16.   </p><div class="table" id="FUNCTIONS-AGGREGATE-TABLE"><p class="title"><strong>Table 9.55. General-Purpose Aggregate Functions</strong></p><div class="table-contents"><table class="table" summary="General-Purpose Aggregate Functions" border="1"><colgroup><col /><col /><col /><col /><col /></colgroup><thead><tr><th>Function</th><th>Argument Type(s)</th><th>Return Type</th><th>Partial Mode</th><th>Description</th></tr></thead><tbody><tr><td>

  17.        <a id="id-1.5.8.25.4.2.2.1.1.1" class="indexterm"></a>

  18.        <code class="function">array_agg(<em class="replaceable"><code>expression</code></em>)</code>

  19.       </td><td>

  20.        any non-array type

  21.       </td><td>

  22.        array of the argument type

  23.       </td><td>No</td><td>input values, including nulls, concatenated into an array</td></tr><tr><td>

  24.        <code class="function">array_agg(<em class="replaceable"><code>expression</code></em>)</code>

  25.       </td><td>

  26.        any array type

  27.       </td><td>

  28.        same as argument data type

  29.       </td><td>No</td><td>input arrays concatenated into array of one higher dimension

  30.        (inputs must all have same dimensionality,

  31.         and cannot be empty or null)</td></tr><tr><td>

  32.        <a id="id-1.5.8.25.4.2.2.3.1.1" class="indexterm"></a>

  33.        <a id="id-1.5.8.25.4.2.2.3.1.2" class="indexterm"></a>

  34.        <code class="function">avg(<em class="replaceable"><code>expression</code></em>)</code>

  35.       </td><td>

  36.        <code class="type">smallint</code>, <code class="type">int</code>,

  37.        <code class="type">bigint</code>, <code class="type">real</code>, <code class="type">double

  38.        precision</code>, <code class="type">numeric</code>, or <code class="type">interval</code>

  39.       </td><td>

  40.        <code class="type">numeric</code> for any integer-type argument,

  41.        <code class="type">double precision</code> for a floating-point argument,

  42.        otherwise the same as the argument data type

  43.       </td><td>Yes</td><td>the average (arithmetic mean) of all non-null input values</td></tr><tr><td>

  44.        <a id="id-1.5.8.25.4.2.2.4.1.1" class="indexterm"></a>

  45.        <code class="function">bit_and(<em class="replaceable"><code>expression</code></em>)</code>

  46.       </td><td>

  47.        <code class="type">smallint</code>, <code class="type">int</code>, <code class="type">bigint</code>, or

  48.        <code class="type">bit</code>

  49.       </td><td>

  50.         same as argument data type

  51.       </td><td>Yes</td><td>the bitwise AND of all non-null input values, or null if none</td></tr><tr><td>

  52.        <a id="id-1.5.8.25.4.2.2.5.1.1" class="indexterm"></a>

  53.        <code class="function">bit_or(<em class="replaceable"><code>expression</code></em>)</code>

  54.       </td><td>

  55.        <code class="type">smallint</code>, <code class="type">int</code>, <code class="type">bigint</code>, or

  56.        <code class="type">bit</code>

  57.       </td><td>

  58.         same as argument data type

  59.       </td><td>Yes</td><td>the bitwise OR of all non-null input values, or null if none</td></tr><tr><td>

  60.        <a id="id-1.5.8.25.4.2.2.6.1.1" class="indexterm"></a>

  61.        <code class="function">bool_and(<em class="replaceable"><code>expression</code></em>)</code>

  62.       </td><td>

  63.        <code class="type">bool</code>

  64.       </td><td>

  65.        <code class="type">bool</code>

  66.       </td><td>Yes</td><td>true if all input values are true, otherwise false</td></tr><tr><td>

  67.        <a id="id-1.5.8.25.4.2.2.7.1.1" class="indexterm"></a>

  68.        <code class="function">bool_or(<em class="replaceable"><code>expression</code></em>)</code>

  69.       </td><td>

  70.        <code class="type">bool</code>

  71.       </td><td>

  72.        <code class="type">bool</code>

  73.       </td><td>Yes</td><td>true if at least one input value is true, otherwise false</td></tr><tr><td>

  74.        <a id="id-1.5.8.25.4.2.2.8.1.1" class="indexterm"></a>

  75.        <code class="function">count(*)</code>

  76.       </td><td> </td><td><code class="type">bigint</code></td><td>Yes</td><td>number of input rows</td></tr><tr><td><code class="function">count(<em class="replaceable"><code>expression</code></em>)</code></td><td>any</td><td><code class="type">bigint</code></td><td>Yes</td><td>

  77.        number of input rows for which the value of <em class="replaceable"><code>expression</code></em> is not null

  78.       </td></tr><tr><td>

  79.        <a id="id-1.5.8.25.4.2.2.10.1.1" class="indexterm"></a>

  80.        <code class="function">every(<em class="replaceable"><code>expression</code></em>)</code>

  81.       </td><td>

  82.        <code class="type">bool</code>

  83.       </td><td>

  84.        <code class="type">bool</code>

  85.       </td><td>Yes</td><td>equivalent to <code class="function">bool_and</code></td></tr><tr><td>

  86.        <a id="id-1.5.8.25.4.2.2.11.1.1" class="indexterm"></a>

  87.        <code class="function">json_agg(<em class="replaceable"><code>expression</code></em>)</code>

  88.       </td><td>

  89.        <code class="type">any</code>

  90.       </td><td>

  91.        <code class="type">json</code>

  92.       </td><td>No</td><td>aggregates values, including nulls, as a JSON array</td></tr><tr><td>

  93.        <a id="id-1.5.8.25.4.2.2.12.1.1" class="indexterm"></a>

  94.        <code class="function">jsonb_agg(<em class="replaceable"><code>expression</code></em>)</code>

  95.       </td><td>

  96.        <code class="type">any</code>

  97.       </td><td>

  98.        <code class="type">jsonb</code>

  99.       </td><td>No</td><td>aggregates values, including nulls, as a JSON array</td></tr><tr><td>

  100.        <a id="id-1.5.8.25.4.2.2.13.1.1" class="indexterm"></a>

  101.        <code class="function">json_object_agg(<em class="replaceable"><code>name</code></em>, <em class="replaceable"><code>value</code></em>)</code>

  102.       </td><td>

  103.        <code class="type">(any, any)</code>

  104.       </td><td>

  105.        <code class="type">json</code>

  106.       </td><td>No</td><td>aggregates name/value pairs as a JSON object; values can be

  107.        null, but not names</td></tr><tr><td>

  108.        <a id="id-1.5.8.25.4.2.2.14.1.1" class="indexterm"></a>

  109.        <code class="function">jsonb_object_agg(<em class="replaceable"><code>name</code></em>, <em class="replaceable"><code>value</code></em>)</code>

  110.       </td><td>

  111.        <code class="type">(any, any)</code>

  112.       </td><td>

  113.        <code class="type">jsonb</code>

  114.       </td><td>No</td><td>aggregates name/value pairs as a JSON object; values can be

  115.        null, but not names</td></tr><tr><td>

  116.        <a id="id-1.5.8.25.4.2.2.15.1.1" class="indexterm"></a>

  117.        <code class="function">max(<em class="replaceable"><code>expression</code></em>)</code>

  118.       </td><td>any numeric, string, date/time, network, or enum type,

  119.              or arrays of these types</td><td>same as argument type</td><td>Yes</td><td>

  120.        maximum value of <em class="replaceable"><code>expression</code></em> across all non-null input

  121.        values

  122.       </td></tr><tr><td>

  123.        <a id="id-1.5.8.25.4.2.2.16.1.1" class="indexterm"></a>

  124.        <code class="function">min(<em class="replaceable"><code>expression</code></em>)</code>

  125.       </td><td>any numeric, string, date/time, network, or enum type,

  126.              or arrays of these types</td><td>same as argument type</td><td>Yes</td><td>

  127.        minimum value of <em class="replaceable"><code>expression</code></em> across all non-null input

  128.        values

  129.       </td></tr><tr><td>

  130.        <a id="id-1.5.8.25.4.2.2.17.1.1" class="indexterm"></a>

  131.        <code class="function">

  132.          string_agg(<em class="replaceable"><code>expression</code></em>,

  133.                     <em class="replaceable"><code>delimiter</code></em>)

  134.        </code>

  135.       </td><td>

  136.        (<code class="type">text</code>, <code class="type">text</code>) or (<code class="type">bytea</code>, <code class="type">bytea</code>)

  137.       </td><td>

  138.        same as argument types

  139.       </td><td>No</td><td>non-null input values concatenated into a string, separated by delimiter</td></tr><tr><td>

  140.        <a id="id-1.5.8.25.4.2.2.18.1.1" class="indexterm"></a>

  141.        <code class="function">sum(<em class="replaceable"><code>expression</code></em>)</code>

  142.       </td><td>

  143.        <code class="type">smallint</code>, <code class="type">int</code>,

  144.        <code class="type">bigint</code>, <code class="type">real</code>, <code class="type">double

  145.        precision</code>, <code class="type">numeric</code>,

  146.        <code class="type">interval</code>, or <code class="type">money</code>

  147.       </td><td>

  148.        <code class="type">bigint</code> for <code class="type">smallint</code> or

  149.        <code class="type">int</code> arguments, <code class="type">numeric</code> for

  150.        <code class="type">bigint</code> arguments, otherwise the same as the

  151.        argument data type

  152.       </td><td>Yes</td><td>sum of <em class="replaceable"><code>expression</code></em>

  153.        across all non-null input values</td></tr><tr><td>

  154.        <a id="id-1.5.8.25.4.2.2.19.1.1" class="indexterm"></a>

  155.        <code class="function">xmlagg(<em class="replaceable"><code>expression</code></em>)</code>

  156.       </td><td>

  157.        <code class="type">xml</code>

  158.       </td><td>

  159.        <code class="type">xml</code>

  160.       </td><td>No</td><td>concatenation of non-null XML values

  161.        (see also <a class="xref" href="functions-xml.html#FUNCTIONS-XML-XMLAGG" title="9.14.1.7. xmlagg">Section 9.14.1.7</a>)</td></tr></tbody></table></div></div><br class="table-break" /><p>

  162.    It should be noted that except for <code class="function">count</code>,

  163.    these functions return a null value when no rows are selected.  In

  164.    particular, <code class="function">sum</code> of no rows returns null, not

  165.    zero as one might expect, and <code class="function">array_agg</code>

  166.    returns null rather than an empty array when there are no input

  167.    rows.  The <code class="function">coalesce</code> function can be used to

  168.    substitute zero or an empty array for null when necessary.

  169.   </p><p>

  170.    Aggregate functions which support <em class="firstterm">Partial Mode</em>

  171.    are eligible to participate in various optimizations, such as parallel

  172.    aggregation.

  173.   </p><div class="note"><h3 class="title">Note</h3><a id="id-1.5.8.25.7.1" class="indexterm"></a><a id="id-1.5.8.25.7.2" class="indexterm"></a><p>

  174.       Boolean aggregates <code class="function">bool_and</code> and

  175.       <code class="function">bool_or</code> correspond to standard SQL aggregates

  176.       <code class="function">every</code> and <code class="function">any</code> or

  177.       <code class="function">some</code>.

  178.       As for <code class="function">any</code> and <code class="function">some</code>,

  179.       it seems that there is an ambiguity built into the standard syntax:

  180. </p><pre class="programlisting">

  181. SELECT b1 = ANY((SELECT b2 FROM t2 ...)) FROM t1 ...;

  182. </pre><p>

  183.       Here <code class="function">ANY</code> can be considered either as introducing

  184.       a subquery, or as being an aggregate function, if the subquery

  185.       returns one row with a Boolean value.

  186.       Thus the standard name cannot be given to these aggregates.

  187.     </p></div><div class="note"><h3 class="title">Note</h3><p>

  188.     Users accustomed to working with other SQL database management

  189.     systems might be disappointed by the performance of the

  190.     <code class="function">count</code> aggregate when it is applied to the

  191.     entire table. A query like:

  192. </p><pre class="programlisting">

  193. SELECT count(*) FROM sometable;

  194. </pre><p>

  195.     will require effort proportional to the size of the table:

  196.     <span class="productname">PostgreSQL</span> will need to scan either the

  197.     entire table or the entirety of an index which includes all rows in

  198.     the table.

  199.    </p></div><p>

  200.    The aggregate functions <code class="function">array_agg</code>,

  201.    <code class="function">json_agg</code>, <code class="function">jsonb_agg</code>,

  202.    <code class="function">json_object_agg</code>, <code class="function">jsonb_object_agg</code>,

  203.    <code class="function">string_agg</code>,

  204.    and <code class="function">xmlagg</code>, as well as similar user-defined

  205.    aggregate functions, produce meaningfully different result values

  206.    depending on the order of the input values.  This ordering is

  207.    unspecified by default, but can be controlled by writing an

  208.    <code class="literal">ORDER BY</code> clause within the aggregate call, as shown in

  209.    <a class="xref" href="sql-expressions.html#SYNTAX-AGGREGATES" title="4.2.7. Aggregate Expressions">Section 4.2.7</a>.

  210.    Alternatively, supplying the input values from a sorted subquery

  211.    will usually work.  For example:

  212. 

  213. </p><pre class="screen">

  214. SELECT xmlagg(x) FROM (SELECT x FROM test ORDER BY y DESC) AS tab;

  215. </pre><p>

  216. 

  217.    Beware that this approach can fail if the outer query level contains

  218.    additional processing, such as a join, because that might cause the

  219.    subquery's output to be reordered before the aggregate is computed.

  220.   </p><p>

  221.    <a class="xref" href="functions-aggregate.html#FUNCTIONS-AGGREGATE-STATISTICS-TABLE" title="Table 9.56. Aggregate Functions for Statistics">Table 9.56</a> shows

  222.    aggregate functions typically used in statistical analysis.

  223.    (These are separated out merely to avoid cluttering the listing

  224.    of more-commonly-used aggregates.)  Where the description mentions

  225.    <em class="replaceable"><code>N</code></em>, it means the

  226.    number of input rows for which all the input expressions are non-null.

  227.    In all cases, null is returned if the computation is meaningless,

  228.    for example when <em class="replaceable"><code>N</code></em> is zero.

  229.   </p><a id="id-1.5.8.25.11" class="indexterm"></a><a id="id-1.5.8.25.12" class="indexterm"></a><div class="table" id="FUNCTIONS-AGGREGATE-STATISTICS-TABLE"><p class="title"><strong>Table 9.56. Aggregate Functions for Statistics</strong></p><div class="table-contents"><table class="table" summary="Aggregate Functions for Statistics" border="1"><colgroup><col /><col /><col /><col /><col /></colgroup><thead><tr><th>Function</th><th>Argument Type</th><th>Return Type</th><th>Partial Mode</th><th>Description</th></tr></thead><tbody><tr><td>

  230.        <a id="id-1.5.8.25.13.2.2.1.1.1" class="indexterm"></a>

  231.        <a id="id-1.5.8.25.13.2.2.1.1.2" class="indexterm"></a>

  232.        <code class="function">corr(<em class="replaceable"><code>Y</code></em>, <em class="replaceable"><code>X</code></em>)</code>

  233.       </td><td>

  234.        <code class="type">double precision</code>

  235.       </td><td>

  236.        <code class="type">double precision</code>

  237.       </td><td>Yes</td><td>correlation coefficient</td></tr><tr><td>

  238.        <a id="id-1.5.8.25.13.2.2.2.1.1" class="indexterm"></a>

  239.        <a id="id-1.5.8.25.13.2.2.2.1.2" class="indexterm"></a>

  240.        <code class="function">covar_pop(<em class="replaceable"><code>Y</code></em>, <em class="replaceable"><code>X</code></em>)</code>

  241.       </td><td>

  242.        <code class="type">double precision</code>

  243.       </td><td>

  244.        <code class="type">double precision</code>

  245.       </td><td>Yes</td><td>population covariance</td></tr><tr><td>

  246.        <a id="id-1.5.8.25.13.2.2.3.1.1" class="indexterm"></a>

  247.        <a id="id-1.5.8.25.13.2.2.3.1.2" class="indexterm"></a>

  248.        <code class="function">covar_samp(<em class="replaceable"><code>Y</code></em>, <em class="replaceable"><code>X</code></em>)</code>

  249.       </td><td>

  250.        <code class="type">double precision</code>

  251.       </td><td>

  252.        <code class="type">double precision</code>

  253.       </td><td>Yes</td><td>sample covariance</td></tr><tr><td>

  254.        <a id="id-1.5.8.25.13.2.2.4.1.1" class="indexterm"></a>

  255.        <code class="function">regr_avgx(<em class="replaceable"><code>Y</code></em>, <em class="replaceable"><code>X</code></em>)</code>

  256.       </td><td>

  257.        <code class="type">double precision</code>

  258.       </td><td>

  259.        <code class="type">double precision</code>

  260.       </td><td>Yes</td><td>average of the independent variable

  261.       (<code class="literal">sum(<em class="replaceable"><code>X</code></em>)/<em class="replaceable"><code>N</code></em></code>)</td></tr><tr><td>

  262.        <a id="id-1.5.8.25.13.2.2.5.1.1" class="indexterm"></a>

  263.        <code class="function">regr_avgy(<em class="replaceable"><code>Y</code></em>, <em class="replaceable"><code>X</code></em>)</code>

  264.       </td><td>

  265.        <code class="type">double precision</code>

  266.       </td><td>

  267.        <code class="type">double precision</code>

  268.       </td><td>Yes</td><td>average of the dependent variable

  269.       (<code class="literal">sum(<em class="replaceable"><code>Y</code></em>)/<em class="replaceable"><code>N</code></em></code>)</td></tr><tr><td>

  270.        <a id="id-1.5.8.25.13.2.2.6.1.1" class="indexterm"></a>

  271.        <code class="function">regr_count(<em class="replaceable"><code>Y</code></em>, <em class="replaceable"><code>X</code></em>)</code>

  272.       </td><td>

  273.        <code class="type">double precision</code>

  274.       </td><td>

  275.        <code class="type">bigint</code>

  276.       </td><td>Yes</td><td>number of input rows in which both expressions are nonnull</td></tr><tr><td>

  277.        <a id="id-1.5.8.25.13.2.2.7.1.1" class="indexterm"></a>

  278.        <a id="id-1.5.8.25.13.2.2.7.1.2" class="indexterm"></a>

  279.        <code class="function">regr_intercept(<em class="replaceable"><code>Y</code></em>, <em class="replaceable"><code>X</code></em>)</code>

  280.       </td><td>

  281.        <code class="type">double precision</code>

  282.       </td><td>

  283.        <code class="type">double precision</code>

  284.       </td><td>Yes</td><td>y-intercept of the least-squares-fit linear equation

  285.       determined by the (<em class="replaceable"><code>X</code></em>, <em class="replaceable"><code>Y</code></em>) pairs</td></tr><tr><td>

  286.        <a id="id-1.5.8.25.13.2.2.8.1.1" class="indexterm"></a>

  287.        <code class="function">regr_r2(<em class="replaceable"><code>Y</code></em>, <em class="replaceable"><code>X</code></em>)</code>

  288.       </td><td>

  289.        <code class="type">double precision</code>

  290.       </td><td>

  291.        <code class="type">double precision</code>

  292.       </td><td>Yes</td><td>square of the correlation coefficient</td></tr><tr><td>

  293.        <a id="id-1.5.8.25.13.2.2.9.1.1" class="indexterm"></a>

  294.        <a id="id-1.5.8.25.13.2.2.9.1.2" class="indexterm"></a>

  295.        <code class="function">regr_slope(<em class="replaceable"><code>Y</code></em>, <em class="replaceable"><code>X</code></em>)</code>

  296.       </td><td>

  297.        <code class="type">double precision</code>

  298.       </td><td>

  299.        <code class="type">double precision</code>

  300.       </td><td>Yes</td><td>slope of the least-squares-fit linear equation determined

  301.       by the (<em class="replaceable"><code>X</code></em>,

  302.       <em class="replaceable"><code>Y</code></em>) pairs</td></tr><tr><td>

  303.        <a id="id-1.5.8.25.13.2.2.10.1.1" class="indexterm"></a>

  304.        <code class="function">regr_sxx(<em class="replaceable"><code>Y</code></em>, <em class="replaceable"><code>X</code></em>)</code>

  305.       </td><td>

  306.        <code class="type">double precision</code>

  307.       </td><td>

  308.        <code class="type">double precision</code>

  309.       </td><td>Yes</td><td><code class="literal">sum(<em class="replaceable"><code>X</code></em>^2) - sum(<em class="replaceable"><code>X</code></em>)^2/<em class="replaceable"><code>N</code></em></code> (<span class="quote">“<span class="quote">sum of

  310.       squares</span>”</span> of the independent variable)</td></tr><tr><td>

  311.        <a id="id-1.5.8.25.13.2.2.11.1.1" class="indexterm"></a>

  312.        <code class="function">regr_sxy(<em class="replaceable"><code>Y</code></em>, <em class="replaceable"><code>X</code></em>)</code>

  313.       </td><td>

  314.        <code class="type">double precision</code>

  315.       </td><td>

  316.        <code class="type">double precision</code>

  317.       </td><td>Yes</td><td><code class="literal">sum(<em class="replaceable"><code>X</code></em>*<em class="replaceable"><code>Y</code></em>) - sum(<em class="replaceable"><code>X</code></em>) * sum(<em class="replaceable"><code>Y</code></em>)/<em class="replaceable"><code>N</code></em></code> (<span class="quote">“<span class="quote">sum of

  318.       products</span>”</span> of independent times dependent

  319.       variable)</td></tr><tr><td>

  320.        <a id="id-1.5.8.25.13.2.2.12.1.1" class="indexterm"></a>

  321.        <code class="function">regr_syy(<em class="replaceable"><code>Y</code></em>, <em class="replaceable"><code>X</code></em>)</code>

  322.       </td><td>

  323.        <code class="type">double precision</code>

  324.       </td><td>

  325.        <code class="type">double precision</code>

  326.       </td><td>Yes</td><td><code class="literal">sum(<em class="replaceable"><code>Y</code></em>^2) - sum(<em class="replaceable"><code>Y</code></em>)^2/<em class="replaceable"><code>N</code></em></code> (<span class="quote">“<span class="quote">sum of

  327.       squares</span>”</span> of the dependent variable)</td></tr><tr><td>

  328.        <a id="id-1.5.8.25.13.2.2.13.1.1" class="indexterm"></a>

  329.        <a id="id-1.5.8.25.13.2.2.13.1.2" class="indexterm"></a>

  330.        <code class="function">stddev(<em class="replaceable"><code>expression</code></em>)</code>

  331.       </td><td>

  332.        <code class="type">smallint</code>, <code class="type">int</code>,

  333.        <code class="type">bigint</code>, <code class="type">real</code>, <code class="type">double

  334.        precision</code>, or <code class="type">numeric</code>

  335.       </td><td>

  336.        <code class="type">double precision</code> for floating-point arguments,

  337.        otherwise <code class="type">numeric</code>

  338.       </td><td>Yes</td><td>historical alias for <code class="function">stddev_samp</code></td></tr><tr><td>

  339.        <a id="id-1.5.8.25.13.2.2.14.1.1" class="indexterm"></a>

  340.        <a id="id-1.5.8.25.13.2.2.14.1.2" class="indexterm"></a>

  341.        <code class="function">stddev_pop(<em class="replaceable"><code>expression</code></em>)</code>

  342.       </td><td>

  343.        <code class="type">smallint</code>, <code class="type">int</code>,

  344.        <code class="type">bigint</code>, <code class="type">real</code>, <code class="type">double

  345.        precision</code>, or <code class="type">numeric</code>

  346.       </td><td>

  347.        <code class="type">double precision</code> for floating-point arguments,

  348.        otherwise <code class="type">numeric</code>

  349.       </td><td>Yes</td><td>population standard deviation of the input values</td></tr><tr><td>

  350.        <a id="id-1.5.8.25.13.2.2.15.1.1" class="indexterm"></a>

  351.        <a id="id-1.5.8.25.13.2.2.15.1.2" class="indexterm"></a>

  352.        <code class="function">stddev_samp(<em class="replaceable"><code>expression</code></em>)</code>

  353.       </td><td>

  354.        <code class="type">smallint</code>, <code class="type">int</code>,

  355.        <code class="type">bigint</code>, <code class="type">real</code>, <code class="type">double

  356.        precision</code>, or <code class="type">numeric</code>

  357.       </td><td>

  358.        <code class="type">double precision</code> for floating-point arguments,

  359.        otherwise <code class="type">numeric</code>

  360.       </td><td>Yes</td><td>sample standard deviation of the input values</td></tr><tr><td>

  361.        <a id="id-1.5.8.25.13.2.2.16.1.1" class="indexterm"></a>

  362.        <code class="function">variance</code>(<em class="replaceable"><code>expression</code></em>)

  363.       </td><td>

  364.        <code class="type">smallint</code>, <code class="type">int</code>,

  365.        <code class="type">bigint</code>, <code class="type">real</code>, <code class="type">double

  366.        precision</code>, or <code class="type">numeric</code>

  367.       </td><td>

  368.        <code class="type">double precision</code> for floating-point arguments,

  369.        otherwise <code class="type">numeric</code>

  370.       </td><td>Yes</td><td>historical alias for <code class="function">var_samp</code></td></tr><tr><td>

  371.        <a id="id-1.5.8.25.13.2.2.17.1.1" class="indexterm"></a>

  372.        <a id="id-1.5.8.25.13.2.2.17.1.2" class="indexterm"></a>

  373.        <code class="function">var_pop</code>(<em class="replaceable"><code>expression</code></em>)

  374.       </td><td>

  375.        <code class="type">smallint</code>, <code class="type">int</code>,

  376.        <code class="type">bigint</code>, <code class="type">real</code>, <code class="type">double

  377.        precision</code>, or <code class="type">numeric</code>

  378.       </td><td>

  379.        <code class="type">double precision</code> for floating-point arguments,

  380.        otherwise <code class="type">numeric</code>

  381.       </td><td>Yes</td><td>population variance of the input values (square of the population standard deviation)</td></tr><tr><td>

  382.        <a id="id-1.5.8.25.13.2.2.18.1.1" class="indexterm"></a>

  383.        <a id="id-1.5.8.25.13.2.2.18.1.2" class="indexterm"></a>

  384.        <code class="function">var_samp</code>(<em class="replaceable"><code>expression</code></em>)

  385.       </td><td>

  386.        <code class="type">smallint</code>, <code class="type">int</code>,

  387.        <code class="type">bigint</code>, <code class="type">real</code>, <code class="type">double

  388.        precision</code>, or <code class="type">numeric</code>

  389.       </td><td>

  390.        <code class="type">double precision</code> for floating-point arguments,

  391.        otherwise <code class="type">numeric</code>

  392.       </td><td>Yes</td><td>sample variance of the input values (square of the sample standard deviation)</td></tr></tbody></table></div></div><br class="table-break" /><p>

  393.    <a class="xref" href="functions-aggregate.html#FUNCTIONS-ORDEREDSET-TABLE" title="Table 9.57. Ordered-Set Aggregate Functions">Table 9.57</a> shows some

  394.    aggregate functions that use the <em class="firstterm">ordered-set aggregate</em>

  395.    syntax.  These functions are sometimes referred to as <span class="quote">“<span class="quote">inverse

  396.    distribution</span>”</span> functions.

  397.   </p><a id="id-1.5.8.25.15" class="indexterm"></a><a id="id-1.5.8.25.16" class="indexterm"></a><div class="table" id="FUNCTIONS-ORDEREDSET-TABLE"><p class="title"><strong>Table 9.57. Ordered-Set Aggregate Functions</strong></p><div class="table-contents"><table class="table" summary="Ordered-Set Aggregate Functions" border="1"><colgroup><col /><col /><col /><col /><col /><col /></colgroup><thead><tr><th>Function</th><th>Direct Argument Type(s)</th><th>Aggregated Argument Type(s)</th><th>Return Type</th><th>Partial Mode</th><th>Description</th></tr></thead><tbody><tr><td>

  398.        <a id="id-1.5.8.25.17.2.2.1.1.1" class="indexterm"></a>

  399.        <code class="function">mode() WITHIN GROUP (ORDER BY <em class="replaceable"><code>sort_expression</code></em>)</code>

  400.       </td><td>

  401.       </td><td>

  402.        any sortable type

  403.       </td><td>

  404.        same as sort expression

  405.       </td><td>No</td><td>

  406.        returns the most frequent input value (arbitrarily choosing the first

  407.        one if there are multiple equally-frequent results)

  408.       </td></tr><tr><td>

  409.        <a id="id-1.5.8.25.17.2.2.2.1.1" class="indexterm"></a>

  410.        <code class="function">percentile_cont(<em class="replaceable"><code>fraction</code></em>) WITHIN GROUP (ORDER BY <em class="replaceable"><code>sort_expression</code></em>)</code>

  411.       </td><td>

  412.        <code class="type">double precision</code>

  413.       </td><td>

  414.        <code class="type">double precision</code> or <code class="type">interval</code>

  415.       </td><td>

  416.        same as sort expression

  417.       </td><td>No</td><td>

  418.        continuous percentile: returns a value corresponding to the specified

  419.        fraction in the ordering, interpolating between adjacent input items if

  420.        needed

  421.       </td></tr><tr><td>

  422.        <code class="function">percentile_cont(<em class="replaceable"><code>fractions</code></em>) WITHIN GROUP (ORDER BY <em class="replaceable"><code>sort_expression</code></em>)</code>

  423.       </td><td>

  424.        <code class="type">double precision[]</code>

  425.       </td><td>

  426.        <code class="type">double precision</code> or <code class="type">interval</code>

  427.       </td><td>

  428.        array of sort expression's type

  429.       </td><td>No</td><td>

  430.        multiple continuous percentile: returns an array of results matching

  431.        the shape of the <em class="replaceable"><code>fractions</code></em> parameter, with each

  432.        non-null element replaced by the value corresponding to that percentile

  433.       </td></tr><tr><td>

  434.        <a id="id-1.5.8.25.17.2.2.4.1.1" class="indexterm"></a>

  435.        <code class="function">percentile_disc(<em class="replaceable"><code>fraction</code></em>) WITHIN GROUP (ORDER BY <em class="replaceable"><code>sort_expression</code></em>)</code>

  436.       </td><td>

  437.        <code class="type">double precision</code>

  438.       </td><td>

  439.        any sortable type

  440.       </td><td>

  441.        same as sort expression

  442.       </td><td>No</td><td>

  443.        discrete percentile: returns the first input value whose position in

  444.        the ordering equals or exceeds the specified fraction

  445.       </td></tr><tr><td>

  446.        <code class="function">percentile_disc(<em class="replaceable"><code>fractions</code></em>) WITHIN GROUP (ORDER BY <em class="replaceable"><code>sort_expression</code></em>)</code>

  447.       </td><td>

  448.        <code class="type">double precision[]</code>

  449.       </td><td>

  450.        any sortable type

  451.       </td><td>

  452.        array of sort expression's type

  453.       </td><td>No</td><td>

  454.        multiple discrete percentile: returns an array of results matching the

  455.        shape of the <em class="replaceable"><code>fractions</code></em> parameter, with each non-null

  456.        element replaced by the input value corresponding to that percentile

  457.       </td></tr></tbody></table></div></div><br class="table-break" /><p>

  458.    All the aggregates listed in <a class="xref" href="functions-aggregate.html#FUNCTIONS-ORDEREDSET-TABLE" title="Table 9.57. Ordered-Set Aggregate Functions">Table 9.57</a>

  459.    ignore null values in their sorted input.  For those that take

  460.    a <em class="replaceable"><code>fraction</code></em> parameter, the fraction value must be

  461.    between 0 and 1; an error is thrown if not.  However, a null fraction value

  462.    simply produces a null result.

  463.   </p><a id="id-1.5.8.25.19" class="indexterm"></a><p>

  464.    Each of the aggregates listed in

  465.    <a class="xref" href="functions-aggregate.html#FUNCTIONS-HYPOTHETICAL-TABLE" title="Table 9.58. Hypothetical-Set Aggregate Functions">Table 9.58</a> is associated with a

  466.    window function of the same name defined in

  467.    <a class="xref" href="functions-window.html" title="9.21. Window Functions">Section 9.21</a>.  In each case, the aggregate result

  468.    is the value that the associated window function would have

  469.    returned for the <span class="quote">“<span class="quote">hypothetical</span>”</span> row constructed from

  470.    <em class="replaceable"><code>args</code></em>, if such a row had been added to the sorted

  471.    group of rows computed from the <em class="replaceable"><code>sorted_args</code></em>.

  472.   </p><div class="table" id="FUNCTIONS-HYPOTHETICAL-TABLE"><p class="title"><strong>Table 9.58. Hypothetical-Set Aggregate Functions</strong></p><div class="table-contents"><table class="table" summary="Hypothetical-Set Aggregate Functions" border="1"><colgroup><col /><col /><col /><col /><col /><col /></colgroup><thead><tr><th>Function</th><th>Direct Argument Type(s)</th><th>Aggregated Argument Type(s)</th><th>Return Type</th><th>Partial Mode</th><th>Description</th></tr></thead><tbody><tr><td>

  473.        <a id="id-1.5.8.25.21.2.2.1.1.1" class="indexterm"></a>

  474.        <code class="function">rank(<em class="replaceable"><code>args</code></em>) WITHIN GROUP (ORDER BY <em class="replaceable"><code>sorted_args</code></em>)</code>

  475.       </td><td>

  476.        <code class="literal">VARIADIC</code> <code class="type">"any"</code>

  477.       </td><td>

  478.        <code class="literal">VARIADIC</code> <code class="type">"any"</code>

  479.       </td><td>

  480.        <code class="type">bigint</code>

  481.       </td><td>No</td><td>

  482.        rank of the hypothetical row, with gaps for duplicate rows

  483.       </td></tr><tr><td>

  484.        <a id="id-1.5.8.25.21.2.2.2.1.1" class="indexterm"></a>

  485.        <code class="function">dense_rank(<em class="replaceable"><code>args</code></em>) WITHIN GROUP (ORDER BY <em class="replaceable"><code>sorted_args</code></em>)</code>

  486.       </td><td>

  487.        <code class="literal">VARIADIC</code> <code class="type">"any"</code>

  488.       </td><td>

  489.        <code class="literal">VARIADIC</code> <code class="type">"any"</code>

  490.       </td><td>

  491.        <code class="type">bigint</code>

  492.       </td><td>No</td><td>

  493.        rank of the hypothetical row, without gaps

  494.       </td></tr><tr><td>

  495.        <a id="id-1.5.8.25.21.2.2.3.1.1" class="indexterm"></a>

  496.        <code class="function">percent_rank(<em class="replaceable"><code>args</code></em>) WITHIN GROUP (ORDER BY <em class="replaceable"><code>sorted_args</code></em>)</code>

  497.       </td><td>

  498.        <code class="literal">VARIADIC</code> <code class="type">"any"</code>

  499.       </td><td>

  500.        <code class="literal">VARIADIC</code> <code class="type">"any"</code>

  501.       </td><td>

  502.        <code class="type">double precision</code>

  503.       </td><td>No</td><td>

  504.        relative rank of the hypothetical row, ranging from 0 to 1

  505.       </td></tr><tr><td>

  506.        <a id="id-1.5.8.25.21.2.2.4.1.1" class="indexterm"></a>

  507.        <code class="function">cume_dist(<em class="replaceable"><code>args</code></em>) WITHIN GROUP (ORDER BY <em class="replaceable"><code>sorted_args</code></em>)</code>

  508.       </td><td>

  509.        <code class="literal">VARIADIC</code> <code class="type">"any"</code>

  510.       </td><td>

  511.        <code class="literal">VARIADIC</code> <code class="type">"any"</code>

  512.       </td><td>

  513.        <code class="type">double precision</code>

  514.       </td><td>No</td><td>

  515.        relative rank of the hypothetical row, ranging from

  516.        1/<em class="replaceable"><code>N</code></em> to 1

  517.       </td></tr></tbody></table></div></div><br class="table-break" /><p>

  518.    For each of these hypothetical-set aggregates, the list of direct arguments

  519.    given in <em class="replaceable"><code>args</code></em> must match the number and types of

  520.    the aggregated arguments given in <em class="replaceable"><code>sorted_args</code></em>.

  521.    Unlike most built-in aggregates, these aggregates are not strict, that is

  522.    they do not drop input rows containing nulls.  Null values sort according

  523.    to the rule specified in the <code class="literal">ORDER BY</code> clause.

  524.   </p><div class="table" id="FUNCTIONS-GROUPING-TABLE"><p class="title"><strong>Table 9.59. Grouping Operations</strong></p><div class="table-contents"><table class="table" summary="Grouping Operations" border="1"><colgroup><col /><col /><col /></colgroup><thead><tr><th>Function</th><th>Return Type</th><th>Description</th></tr></thead><tbody><tr><td>

  525.        <a id="id-1.5.8.25.23.2.2.1.1.1" class="indexterm"></a>

  526.        <code class="function">GROUPING(<em class="replaceable"><code>args...</code></em>)</code>

  527.       </td><td>

  528.        <code class="type">integer</code>

  529.       </td><td>

  530.        Integer bit mask indicating which arguments are not being included in the current

  531.        grouping set

  532.       </td></tr></tbody></table></div></div><br class="table-break" /><p>

  533.     Grouping operations are used in conjunction with grouping sets (see

  534.     <a class="xref" href="queries-table-expressions.html#QUERIES-GROUPING-SETS" title="7.2.4. GROUPING SETS, CUBE, and ROLLUP">Section 7.2.4</a>) to distinguish result rows.  The

  535.     arguments to the <code class="literal">GROUPING</code> operation are not actually evaluated,

  536.     but they must match exactly expressions given in the <code class="literal">GROUP BY</code>

  537.     clause of the associated query level.  Bits are assigned with the rightmost

  538.     argument being the least-significant bit; each bit is 0 if the corresponding

  539.     expression is included in the grouping criteria of the grouping set generating

  540.     the result row, and 1 if it is not.  For example:

  541. </p><pre class="screen">

  542. <code class="prompt">=&gt;</code> <strong class="userinput"><code>SELECT * FROM items_sold;</code></strong>

  543.  make  | model | sales

  544. -------+-------+-------

  545.  Foo   | GT    |  10

  546.  Foo   | Tour  |  20

  547.  Bar   | City  |  15

  548.  Bar   | Sport |  5

  549. (4 rows)

  550. 

  551. <code class="prompt">=&gt;</code> <strong class="userinput"><code>SELECT make, model, GROUPING(make,model), sum(sales) FROM items_sold GROUP BY ROLLUP(make,model);</code></strong>

  552.  make  | model | grouping | sum

  553. -------+-------+----------+-----

  554.  Foo   | GT    |        0 | 10

  555.  Foo   | Tour  |        0 | 20

  556.  Bar   | City  |        0 | 15

  557.  Bar   | Sport |        0 | 5

  558.  Foo   |       |        1 | 30

  559.  Bar   |       |        1 | 20

  560.        |       |        3 | 50

  561. (7 rows)

  562. </pre><p>

  563.    </p></div><div class="navfooter"><hr /><table width="100%" summary="Navigation footer"><tr><td width="40%" align="left"><a accesskey="p" href="functions-range.html">Prev</a> </td><td width="20%" align="center"><a accesskey="u" href="functions.html">Up</a></td><td width="40%" align="right"> <a accesskey="n" href="functions-window.html">Next</a></td></tr><tr><td width="40%" align="left" valign="top">9.19. Range Functions and Operators </td><td width="20%" align="center"><a accesskey="h" href="index.html">Home</a></td><td width="40%" align="right" valign="top"> 9.21. Window Functions</td></tr></table></div></body></html>
上海开阖软件有限公司 沪ICP备12045867号-1