tsdiff [ parameter=value ... ] [ infile1 infile2 outfile ]
Parameters are: include_vars, report_only.
tsdiff computes the difference between two datasets for one or more variables. Each specified variable must exist in both input datasets, and have the same dimensions. tsdiff can either store the differences in an output dataset, or simply generate a report. Unfortunately, the output dataset must be specified, even if only the report is desired.
For each differenced variable, the following information is reported: total number of elements, number of elements that were good in both input datasets, number of elements that were identical, the minimum and maximum difference, and the standard deviation of the difference.
The following example computes the difference between two version of the same dataset, one floating point, the other short integer scaled. Because the scale factor use is 100, the scaled data preserves only two digits beyond the decimal place. As expected, the errors are in the range [-.005, .005].
% fakedata float master_file : char(255) ? [] dim_sizes : int ( 3) ? 10 10 dim_names : char( 63) ? [line sample] expression : char(255) ? 1./(i1 + i2) var_name : char( 31) ? data var_units : char( 31) ? [] var_type : char( 15) ? [float] % thetis% emath float short expr_vars : char(255) ? data expression : char(255) ? x1 var_name : char( 31) ? data var_units : char( 31) ? [] var_type : char( 15) ? [float] short offset : real ? [0] scale_factor : real ? [1] 100 % % tsdiff float short dummy include_vars : char(255) ? [] report_only : char( 3) ? [no] y Count Good Zero Mean Min Max Std Dev Variable 100 100 4 -0.000147686 -0.005 0.00333334 0.0026291 data
Last Update: $Date: 1998/05/29 20:54:48 $