/usr/share/doc/postgresql-9.1-plr/plr-funcs.html

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
<HTML
><HEAD
><TITLE
>Functions and Arguments</TITLE
><META
NAME="GENERATOR"
CONTENT="Modular DocBook HTML Stylesheet Version 1.79"><LINK
REV="MADE"
HREF="mailto:pgsql-docs@postgresql.org"><LINK
REL="HOME"
TITLE="PL/R User's Guide - R Procedural Language"
HREF="index.html"><LINK
REL="PREVIOUS"
TITLE="Installation"
HREF="plr-install.html"><LINK
REL="NEXT"
TITLE="Passing Data Values"
HREF="plr-data.html"><LINK
REL="STYLESHEET"
TYPE="text/css"
HREF="stylesheet.css"><META
HTTP-EQUIV="Content-Type"
CONTENT="text/html; charset=ISO-8859-1"><META
NAME="creation"
CONTENT="2011-09-10T18:39:12"></HEAD
><BODY
CLASS="CHAPTER"
><DIV
CLASS="NAVHEADER"
><TABLE
SUMMARY="Header navigation table"
WIDTH="100%"
BORDER="0"
CELLPADDING="0"
CELLSPACING="0"
><TR
><TH
COLSPAN="5"
ALIGN="center"
VALIGN="bottom"
><A
HREF="index.html"
>PL/R User's Guide - R Procedural Language</A
></TH
></TR
><TR
><TD
WIDTH="10%"
ALIGN="left"
VALIGN="top"
><A
TITLE="Installation"
HREF="plr-install.html"
ACCESSKEY="P"
>Prev</A
></TD
><TD
WIDTH="10%"
ALIGN="left"
VALIGN="top"
><A
TITLE="Installation"
HREF="plr-install.html"
>Fast Backward</A
></TD
><TD
WIDTH="60%"
ALIGN="center"
VALIGN="bottom"
></TD
><TD
WIDTH="10%"
ALIGN="right"
VALIGN="top"
><A
TITLE="Passing Data Values"
HREF="plr-data.html"
>Fast Forward</A
></TD
><TD
WIDTH="10%"
ALIGN="right"
VALIGN="top"
><A
TITLE="Passing Data Values"
HREF="plr-data.html"
ACCESSKEY="N"
>Next</A
></TD
></TR
></TABLE
><HR
ALIGN="LEFT"
WIDTH="100%"></DIV
><DIV
CLASS="CHAPTER"
><H1
><A
NAME="PLR-FUNCS"
></A
>Chapter 3. Functions and Arguments</H1
><P
>    To create a function in the <SPAN
CLASS="APPLICATION"
>PL/R</SPAN
> language, use standard R
    syntax, but without the enclosing braces or function assignment.
    Instead of <TT
CLASS="LITERAL"
>myfunc &#60;- function(arguments) { function body }</TT
>,
    the body of your PL/R function is just <TT
CLASS="LITERAL"
>function body</TT
>

    </P><PRE
CLASS="PROGRAMLISTING"
>CREATE OR REPLACE FUNCTION <TT
CLASS="REPLACEABLE"
><I
>funcname</I
></TT
> (<TT
CLASS="REPLACEABLE"
><I
>argument-types</I
></TT
>)
RETURNS <TT
CLASS="REPLACEABLE"
><I
>return-type</I
></TT
> AS '
    function body
' LANGUAGE 'plr';
    </PRE
><P>
   </P
><P
>    The body of the function is simply a piece of R script. When the function
    is called, the argument values are passed as variables <TT
CLASS="LITERAL"
>    arg1</TT
> ... <TT
CLASS="LITERAL"
>argN</TT
> to the R script.  The result
    is returned from the R code in the usual way. For example, a function
    returning the greater of two integer values could be defined as:

    </P><PRE
CLASS="PROGRAMLISTING"
>CREATE OR REPLACE FUNCTION r_max (integer, integer) RETURNS integer AS '
    if (arg1 &#62; arg2)
       return(arg1)
    else
       return(arg2)
' LANGUAGE 'plr' STRICT;
    </PRE
><P>

    Starting with PostgreSQL 8.0, arguments may be explicitly named when
    creating a function. If an argument is explicitly named at function
    creation time, that name will be available to your R script in place
    of the usual <TT
CLASS="LITERAL"
>argN</TT
> variable. For example:

    </P><PRE
CLASS="PROGRAMLISTING"
>CREATE OR REPLACE FUNCTION sd(vals float8[]) RETURNS float AS '
    sd(vals)
' LANGUAGE 'plr' STRICT;
    </PRE
><P>

   </P
><P
>    Starting with PostgreSQL 8.4, a PL/R function may be declared to be a
    <TT
CLASS="LITERAL"
>WINDOW</TT
>. In this case, in addition to the usual
    <TT
CLASS="LITERAL"
>argN</TT
> (or named) variables, PL/R automatically
    creates several other arguments to your function. For each explicit
    argument, a corresponding variable called <TT
CLASS="LITERAL"
>farg1</TT
>
    ... <TT
CLASS="LITERAL"
>fargN</TT
> is passed to the R script. These contain
    an R vector of all the values of the related argument for the moving
    <TT
CLASS="LITERAL"
>WINDOW</TT
> frame within the current <TT
CLASS="LITERAL"
>PARTITION</TT
>.
    For example:

    </P><PRE
CLASS="PROGRAMLISTING"
>CREATE OR REPLACE
FUNCTION r_regr_slope(float8, float8)
RETURNS float8 AS
$BODY$
  slope &#60;- NA
  y &#60;- farg1
  x &#60;- farg2 
  if (fnumrows==9) try (slope &#60;- lm(y ~ x)$coefficients[2])
  return(slope)
$BODY$
LANGUAGE plr WINDOW;
    </PRE
><P>

    In the preceding example, <TT
CLASS="LITERAL"
>farg1</TT
> and <TT
CLASS="LITERAL"
>farg2</TT
> are
    R vectors containing the current row's data plus that of related
    rows. The determination as to which rows qualify as related is
    determined by the frame specification of the query at run time.
   </P
><P
>    The example also illustrates one of two additional autogenerated
    arguments. <TT
CLASS="LITERAL"
>fnumrows</TT
> is the number of rows in the current
    <TT
CLASS="LITERAL"
>WINDOW</TT
> frame. The other (not shown) auto-argument is called
    <TT
CLASS="LITERAL"
>prownum</TT
>. This argument provides the 1-based row offset of the
    current row in the current <TT
CLASS="LITERAL"
>PARTITION</TT
>.
    See <A
HREF="plr-window-funcs.html"
>Chapter 9</A
> for more information and a more
    complete example.
   </P
><P
>    In some of the the definitions above, note the clause <TT
CLASS="LITERAL"
>STRICT</TT
>,
    which saves us from having to think about NULL input values: if a NULL is
    passed, the function will not be called at all, but will just return a NULL
    result automatically. In a non-strict function, if the actual value of an
    argument is NULL, the corresponding <TT
CLASS="LITERAL"
>argN</TT
> variable will
    be set to a <TT
CLASS="LITERAL"
>NULL</TT
> R object. For example, suppose that we
    wanted <CODE
CLASS="FUNCTION"
>r_max</CODE
> with one null and one non-null argument
    to return the non-null argument, rather than NULL:

    </P><PRE
CLASS="PROGRAMLISTING"
>CREATE OR REPLACE FUNCTION r_max (integer, integer) RETURNS integer AS '
    if (is.null(arg1) &#38;&#38; is.null(arg2))
        return(NULL)
    if (is.null(arg1))
        return(arg2)
    if (is.null(arg2))
        return(arg1)
    if (arg1 &#62; arg2)
       return(arg1)
    arg2
' LANGUAGE 'plr';
    </PRE
><P>
   </P
><P
>    As shown above, to return a NULL value from a PL/R function, return
    <TT
CLASS="LITERAL"
>NULL</TT
>.  This can be done whether the function is strict
    or not.
   </P
><P
>    Composite-type (tuple) arguments are passed to the procedure as
    R data.frames. The element names of the frame are the attribute
    names of the composite type. If an attribute in the passed row
    has the NULL value, it will appear as an "NA" in the frame. Here is
    an example:

    </P><PRE
CLASS="PROGRAMLISTING"
>CREATE TABLE emp (name text, age int, salary numeric(10,2));
INSERT INTO emp VALUES ('Joe', 41, 250000.00);
INSERT INTO emp VALUES ('Jim', 25, 120000.00);
INSERT INTO emp VALUES ('Jon', 35, 50000.00);
CREATE OR REPLACE FUNCTION overpaid (emp) RETURNS bool AS '
    if (200000 &#60; arg1$salary) {
        return(TRUE)
    }
    if (arg1$age &#60; 30 &#38;&#38; 100000 &#60; arg1$salary) {
        return(TRUE)
    }
    return(FALSE)
' LANGUAGE 'plr';
SELECT name, overpaid(emp) FROM emp;
 name | overpaid
------+----------
 Joe  | t
 Jim  | t
 Jon  | f
(3 rows)

    </PRE
><P>
   </P
><P
>    There is also support for returning a composite-type result value:

    </P><PRE
CLASS="PROGRAMLISTING"
>CREATE OR REPLACE FUNCTION get_emps() RETURNS SETOF emp AS '
    names &#60;- c("Joe","Jim","Jon")
    ages &#60;- c(41,25,35)
    salaries &#60;- c(250000,120000,50000)
    df &#60;- data.frame(name = names, age = ages, salary = salaries)
    return(df)
' LANGUAGE 'plr';
select * from get_emps();
 name | age |  salary
------+-----+-----------
 Jim  |  41 | 250000.00
 Joe  |  25 | 120000.00
 Jon  |  35 |  50000.00
(3 rows)

    </PRE
><P>
   </P
><P
>    An alternative method may be used to create a function in PL/R, if
    certain criteria are met. First, the function must be a simple call
    to an existing R function. Second, the function name used for the PL/R
    function must match that of the R function exactly. If these two criteria
    are met, the PL/R function may be defined with no body, and the arguments
    will be passed directly to the R function of the same name. For example:
    </P><PRE
CLASS="PROGRAMLISTING"
>create or replace function sd(_float8) returns float as '' language 'plr';
select round(sd('{1.23,1.31,1.42,1.27}'::_float8)::numeric,8);
   round    
------------
 0.08180261
(1 row)
    </PRE
><P>
   </P
><DIV
CLASS="TIP"
><BLOCKQUOTE
CLASS="TIP"
><P
><B
>Tip: </B
>     Because the function body is passed as an SQL string literal to
     <TT
CLASS="COMMAND"
>CREATE FUNCTION</TT
>, you have to escape single
     quotes and backslashes within your R source, typically by
     doubling them.
    </P
></BLOCKQUOTE
></DIV
></DIV
><DIV
CLASS="NAVFOOTER"
><HR
ALIGN="LEFT"
WIDTH="100%"><TABLE
SUMMARY="Footer navigation table"
WIDTH="100%"
BORDER="0"
CELLPADDING="0"
CELLSPACING="0"
><TR
><TD
WIDTH="33%"
ALIGN="left"
VALIGN="top"
><A
HREF="plr-install.html"
ACCESSKEY="P"
>Prev</A
></TD
><TD
WIDTH="34%"
ALIGN="center"
VALIGN="top"
><A
HREF="index.html"
ACCESSKEY="H"
>Home</A
></TD
><TD
WIDTH="33%"
ALIGN="right"
VALIGN="top"
><A
HREF="plr-data.html"
ACCESSKEY="N"
>Next</A
></TD
></TR
><TR
><TD
WIDTH="33%"
ALIGN="left"
VALIGN="top"
>Installation</TD
><TD
WIDTH="34%"
ALIGN="center"
VALIGN="top"
>&nbsp;</TD
><TD
WIDTH="33%"
ALIGN="right"
VALIGN="top"
>Passing Data Values</TD
></TR
></TABLE
></DIV
></BODY
></HTML
>
postgresql-9.1-plr 1:8.3.0.13-1 / usr / share / doc / postgresql-9.1-plr / plr-funcs.html
