ds_isempty

This macro check whether a dataset, or a variable in the dataset, is empty.

%ds_isempty(dsn, var=, _ans_=, lib=WORK);

Arguments

• dsn : a dataset reference;
• var : (option) a string to be checked whether it exists as a non-empty variable/field in dsn; default: var is empty, and the macro tests whether there are any observation in the dataset or not;
• lib : (option) name of the input library; by default: empty, i.e. WORK is used.

Returns

_ans_ : the name of the macro where the result of the test will be stored, e.g.:

• -1 in the cases: *(i)* the dataset does not exist, and *(ii)* var exists and is not defined as a variable of the dataset
• 0 in the cases: *(i)* the dataset is not empty while var is passed as empty, and *(ii)* var exists in the input dataset and contains non-missing observations,
• 1 otherwise, i.e. in the cases: *(i)* var is passed as empty and the dataset is empty, and *(ii)* var exists in the input dataset but contains only missing observations.

Examples

Let us consider the test on test dataset #2:

a
1

then the following test:

%let ans=;
%_dstest2;
%ds_isempty(_dstest2, var=a, _ans_=ans);

returns ans=0, while:

%_dstest1;
%ds_isempty(_dstest1, var=a, _ans_=ans);

returns ans=1 since the variable a is empty in that latter dataset #1. Even simpler example:

%_dstest0;
%ds_isempty(_dstest0, _ans_=ans);

will naturally report: ans=1.

Run macro %_example_ds_isempty for more examples.

Note

Whenever the variable var is passed but does not exist in the dataset (e.g the test var_check(&dsn, &var, lib=&lib) returns 1), the macro returns ans=-1.

Reference

Childress, S. and Welch, B. (2011): "Three easy ways around nonexistent or empty datasets".

See also

%ds_check, %ds_delete, %var_check.