|
Evaluate the variables and recommend lengths based on maximum
values. %varlen (indata = input data set,
round = round the length,
output = output file);
|
Where
|
Is Type...
|
And represents...
|
| indata |
C |
Source data set to be documented. This
is specified with a two level dot notation: libname.dataname.
If only one level is specified, then all the dataset defined in
the specified libname will be documented. You can also specify
individual datasets with two level names with space separation. |
| round |
C (optional) |
The round option will round the maximum
length numeric value. Valid options include:
|
| output |
C (optional) |
Name of the output HTML file that contains
the results. If this were not to be specified then a varlen.html
file is created in the current location.
|
| overflow |
C (optional) |
For variables that are longer than 200
characters, this option will create a new dataset containing
additional "overflow" variables that contain the values
that go beyond 200 characters. The overflow parameter will
accept a two level data specification with the format of:
libname.dataname
This new dataset will contain original dataset with the additional
overflow variables.
|
Details
This utility will evaluate all the
character variables and determine the maximum length based on the
longest data value of the variable. By default the length is rounded up to
the nearest tenth. For example, if its 23, the length will be set to
30. If the round option is set to 20, it will round to the nearest
length plus 10 for buffer. So 23 will be rounded to 40. A none
will set the length to the exact maximum length. It will create a backup
of the original dataset with the name of "backup". For example, if
the data was AE, the backup data will be named AE_BACKUP. The data
will then have a data step applied retaining any label, format, or
informat while updating the length. Example %varlen (indata = inlib.adverse,
output = c:\temp\varlen.html);
%varlen (indata = inlib,
round = 20,
output = c:\temp\all_data.html);
|
