<!--Copyright (C) 1988-2005 by the Institute of Global Environment and Society (IGES). See file COPYRIGHT for more information.--> <html> <title>GrADS Function: const</title> <body bgcolor="e0f0ff" text="#000000"> <h2><b>const()</b></h2> <p> <p> <code>const (<i>expr, value,</i> <-u|-a>)</code> <p> This is a powerful function that allows the user to change the missing values of a variable, set all the non-missing values of a variable to a constant, or set all possible values of a variable (both valid and missing) to a constant. <p> <ul> <code><i>expr</i> </code>a valid GrADS expression <br> <code><i>value</i> </code>a constant, either an integer or floating point value <br> <code>-u </code>all missing data are set to <code><i>value</i></code>; non-missing data are unchanged<br> <code>-a </code>all data are set to <code><i>value</i></code>, both missing and non-missing<br> </ul> <p> Default behaviour is to set all non-missing data equal to <code><i>value</i></code>; missing data are unchanged <br> </ul> <p> <h3>Usage Notes</h3> <ol> <li>The <code>const</code> function operates on both gridded and station data. <li>If <code><i>value</i></code> is given as an integer, it will still be treated as as floating point. </ol> <p> <h3>Examples</h3><p> <ol> <li>The <code>const</code> function assigns a new value to missing data, so that missing data may participate in operations: <p> <ul><code>const(z, 0, -u)</code></ul> <p> <li>The <code>const</code> function is used with the <a href="gradcomdsetgxout.html"<code>set gxout linefill</code></a> graphics output option to define a straight horizontal line: <p> <ul> <code> set lon -90 <br> set lat -90 90 <br> set gxout linefill <br> set lev 500 <br> d const(t, -20);t-273 <br> </code> </ul> <p> <li>In this example, <code>const</code> is used to calculate a daily timeseries of the fraction of the globe convered by precipitation greater than <code>10mm/day</code>: <p> <ul> <code> set lon 0 360 <br> set lat -90 90 <br> set t 1 last <br> define ones = const(const(maskout(p,p-10),1),0,-u) <br> set x 1 <br> set y 1 <br> display tloop(aave(ones,lon=0,lon=360,lat=0,lat=360)) <br> </code> </ul> <p> Notes: The <a href="gradcomddefine.html">defined</a> variable <code>"ones"</code> contains <code>1</code> wherever the precip value is greater than <code>10</code>, and <code>0</code> whever the precip value is less than <code>10</code>. This is done via nested functions; first <a href="gradfuncmaskout.html"><code>maskout</code></a> sets all values less than <code>10</code> to missing, then <code>const</code> sets all non-missing values to <code>1</code>, then <code>const</code> is used with the <code>-u</code> flag to set all the missing data values to <code>0</code>. The <a href="gradfuncaave.html"><code>aave</code></a> function calculates an area weighted average. Since we are averaging zeros and ones, the result is the fraction of the area where there are ones. See the <a href="gradfunctloop.html"><code>tloop</code></a> function for a description of how to perform time series of areal averages. </ol> </body> </html>