SAS Chapter 22: Using Best Practices

1. For selecting observations, is a subsetting IF statement or a WHERE statement more efficient? Why?

A subsetting IF statement is more efficient because it loads all observations sequentially into the program data vector.

A WHERE statement is more efficient because it loads all observations sequentially into the program data vector.

For selecting observations, a WHERE statement is more efficient than a subsetting IF statement because it examines what is in the input buffer and selects observations before they are loaded into the program data vector, which results in a savings in CPU operations.

Explanation

For selecting observations, a WHERE statement is more efficient than a subsetting IF statement because it examines what is in the input buffer and selects observations before they are loaded into the program data vector, which results in a savings in CPU operations.

2. Which of the following statements is true about techniques that are used for modifying data and attributes?

You can use PROC DATASETS to modify both data values and variable attributes.

You can use PROC DATASETS to modify only data values.

You can use the DATA step to modify both data values and variable attributes.

You can use the DATA step to modify only variable attributes.

The DATA step is the only technique that can be used to modify both data values and variable attributes. The DATASETS procedure enables you to modify only variable attributes.

Explanation

The DATA step is the only technique that can be used to modify both data values and variable attributes. The DATASETS procedure enables you to modify only variable attributes.

3. When is it more advantageous to create a temporary SAS data set rather than a permanent SAS data set?

When the external file on which the data set is based might change between SAS sessions.

When the external file on which the data set is based does not change between SAS sessions.

When the data set is needed for more than one SAS session.

When you are converting raw numeric values to SAS data values.

It is more advantageous to create a temporary SAS data set rather than a permanent SAS data set when the external file on which the data set is based is frequently updated between SAS sessions.

Explanation

It is more advantageous to create a temporary SAS data set rather than a permanent SAS data set when the external file on which the data set is based is frequently updated between SAS sessions.

4. Placing the subsetting IF statement at the top rather than near the bottom of a DATA step results in a savings in CPU usage. What happens if the subset is large rather than small?

The savings in CPU usage increases as the subset grows larger because the I/O increases.

As SAS processes a larger subset of the data, more CPU resources are required. However, positioning of the subsetting IF statement in a DATA step does affect performance and efficiency.

Explanation

As SAS processes a larger subset of the data, more CPU resources are required. However, positioning of the subsetting IF statement in a DATA step does affect performance and efficiency.

5. When you compare the technique of using multiple DATASETS procedures to using a single DATASETS procedure to modify the descriptor portion of a data set, which is true?

A one-step DATASETS procedure results in an increase in I/O operations.

Multiple DATASETS procedures result in a decrease in I/O operations.

A one-step DATASETS procedure results in a decrease in CPU usage.

Multiple DATASETS procedures result in a decrease in CPU usage.

A one-step DATASETS procedure results in a savings of CPU usage and I/O operations. PROC DATASETS supports RUN-group processing, which enables you to process multiple SAS data sets from the same library with one invocation of the procedure.

Explanation

A one-step DATASETS procedure results in a savings of CPU usage and I/O operations. PROC DATASETS supports RUN-group processing, which enables you to process multiple SAS data sets from the same library with one invocation of the procedure.