SAS data _null_ – Create a SAS Dataset with No Records and Variables

In SAS, the _null_ keyword in a SAS Data Step allows us to create a SAS dataset with no records and variables. With _null_, surprisingly, we are able to many things in our SAS code.

In this post, you will learn the many benefits of using _null_ in your SAS code.

Using the SAS _null_ Keyword in Your SAS Code

The SAS keyword _null_ creates a SAS dataset with no observations and no variables.

You can use _null_ in a data step for both the input and output dataset.

If it is specified as the input dataset, then you will create a dataset with no observations and no variables as shown below.

data out;
    set _null_;
run;

Using _null_ in this case can be useful if you want to initialize an empty dataset and you will add records to that dataset in a SAS Macro loop, for example.

data out;
    set _null_;
run;

%macro example_loop;
    %do i = 0 %to 10;

        /* some calculations */

        data out other_dataset;
             set out;
        run;
    %end;
%mend;

Otherwise, if you are using _null_ as the output dataset, then nothing will be created after the data step.

Even though this may seem useless, there are actually many ways you can use data _null_ to perform different tasks in SAS.

The Many Uses of data _null_ in SAS

There are a number of ways you can use data _null_ in a SAS data step.

One very common way you can use data _null_ is to create macro variables.

You can use the symput() and symputx() subroutines to create SAS macro variables from values in a SAS data set.

For example, if you want to get a value from a specific variable and create a macro variable from that value, you can do so as shown below.

data _null_;
    set summary_dataset;
    if name = "Sally" then call symputx("SallyVar", variable);
run;

You can also use data _null_ to put different values to the log, such as the value of a variable after a calculation or information about the dataset.

For example, if you want to do a specific calculation and put that value to the log, then you can use data _null_

data _null_;
    x = 5 * 10;
    put x;
run;

The majority of cases you will use data _null_ for fall into these two categories – creating macro variables or putting information to the log.

Hopefully this article has been useful for you to learn how to use data _null_ in your SAS code.