cdfrdim −− averages over a particular dimension
cdfrdim dimension_name lower_limit [upper_limit]
Cdfrdim accepts a common data format file on the standard input and writes a modified common data format file to the standard output in which data over the dimension specified on the command line are averaged. Fields with the specified dimension are reduced in dimension by one in the output file. The second and third command line arguments give the limits over which the averaging takes place. If the third argument is missing, then it’s value is assumed to be the same as the second argument.
Bad or missing data are omitted from the average. If no good data occur, then the average itself is set to be bad. Data with absolute values greater than the header parameter "badlim" are considered to be bad or missing. If this parameter is absent, the default value 9.99e29 is assumed.
The actual domain over which averaging takes place is determined by matching command line limits with values in the index field of the specified dimension. (Thus, the index field must exist!) If exact correspondence between a limit and an index field entry doesn’t exist, the closest index field entry is chosen as the actual limit.
kestrel%cdfrdim z 5 15 < input_file > output_file
This example would eliminate the z dimension by averaging fields over the interval [5, 15] in z.
kestrel%cdfrdim x −2 < input_file > output_file
In this example −2 is assumed for the upper limit as well as the lower limit, and the averaging process degenerates to extracting the values of fields at x = −2.