<!--Copyright (C) 1988-2005 by the Institute of Global Environment and Society (IGES). See file COPYRIGHT for more information.--> <code>set gxout stat <br> d </code><br> <ul>sends output to the terminal as opposed to a plot or data output (e.g., <code>set fwrite out.dat ; set gxout fwrite; d rh</code>). Or the output goes to the script variable <code>result</code> which can be parsed inside a script (see the <code>corr.gs</code> GrADS script)<p> The output allows many statistical calculations to be made. Here's an example of opening up a global model file and looking at the 1000 mb relative humidity, statistically,<p> <ul> <code>ga-> set gxout stat <br> ga-> d rh <br> Data Type = grid <br> Dimensions = 0 1<br> I Dimension = 1 to 145 <br> J Dimension = 1 to 73 <br> Sizes = 145 73 10585<br> Undef value = 1e+20 <br> Undef count = 0 Valid count = 10585 <br> Min, Max = 0.0610352 100.061 <br> Stats(sum,sumsqr,n): 787381 6.35439e+07 10585 <br> Stats(sum,sumsqr)/n: 74.3865 6003.2<br> Stats(sum,sumsqr)/(n-1): 74.3935 6003.77 <br> Stats(sigma,var)(n): 21.6761 469.854 <br> Stats(sigma,var)(n-1): 21.6771 469.898<br> Cmin, cmax, cint = 10 100 10</code></ul><p> Let's break it down:<p> <ul> <code>Data Type = grid</code> ----- you have a grid <p> <code>Dimensions = 0 1</code> ----- the dimension type for the variable <br> <ul> <code>0</code> - lon <br> <code>1</code> - lat <br> <code>2</code> - lev <br> <code>3</code> - time </ul><br> <code>1</code> - not varying <p> <code>I Dimension = 1 to 145</code> ------ obvious <p> <code>J Dimension = 1 to 73</code><p> <code>Sizes = 145 73 10585</code> ------- <code>10585</code> is 145*73 or total number of points <p> <code>Undef value = 1e+20</code> ------- undefined value <p> <code>Undef count = 0 Valid count = 10585</code> ----- # of defined and undefined points in the grid. Remember that if GrADS can't find any data it returns undefs. This is useful for checking if you have any data, <code>Valid count = 0</code> means no... <p> <code>Min, Max = 0.0610352 100.061</code> ---- UHHH OHHHH! we have slight supersaturation..<p> <code>Stats(sum,sumsqr,n): 787381 6.35439e+07 10585</code> - This should be fairly obvious, sum = the simple sum of all <b>defined</b> points.<p> <code>sumsqr</code> = sum of, in this case, <code>rh*rh</code> and <code>10585</code> is <code>n</code>.<p> <code>Stats(sum,sumsqr)/n: 74.3865 6003.2</code> - Divide by <code>n</code> for convenience, the first number is the "biased" mean...<p> <code>Stats(sum,sumsqr)/(n-1): 74.3935 6003.77</code> - the so called <code>unbiased</code> mean (remove 1 degree of freedom), etc.<p> <code>Stats(sigma,var)(n): 21.6761 469.854</code> - the standard deviation and variance "biased" (<code>n</code>) <p> <code>Stats(sigma,var)(n-1): 21.6771 469.898</code> - the standard deviation and variance "unbiased" (<code>n-1</code>)<p> <code>Cmin, cmax, cint = 10 100 10</code> - What GrADS will use when contouring.<p> </ul><b>NOTE</b>: This works for both <code>gridded</code> and <b>station</b> data