SAS Average Calculation (PROC MEANS) Code Generator

Quickly generate robust SAS code to calculate the mean of your data. This tool helps you create syntax for simple averages, grouped summaries with the CLASS statement, and options for outputting the results to a new dataset.

Generate Your SAS Code

Generated SAS Code

Code copied to clipboard!

Code Explanation

This code uses PROC MEANS to calculate descriptive statistics for the specified variables.

Example SAS Output Table

Type	_STAT_	MSRP	Invoice
Sedan	MEAN	33000.50	30500.75
SUV	MEAN	38500.00	35800.20

This table shows a simplified example of how SAS might display the mean values for each vehicle ‘Type’.

What is Calculating an Average in SAS?

To calculate an average (or mean) in SAS is to compute the central tendency of a numeric variable. This is one of the most fundamental tasks in descriptive statistics and data analysis. While there are several methods, the most common and powerful tool for this is the PROC MEANS procedure. It is specifically designed to generate a wide array of summary statistics, including the mean, for one or more variables.

Beyond a simple overall average, SAS excels at calculating averages for subgroups within your data. For instance, you might not just want the average salary for all employees, but the average salary broken down by department, job title, or region. This is easily accomplished in PROC MEANS using a CLASS statement. Other procedures like PROC SQL can also be used, offering a syntax familiar to database users.

The SAS ‘Formula’: PROC MEANS Syntax Explained

In SAS, the “formula” to calculate an average isn’t a mathematical equation but a syntax structure. The PROC MEANS procedure is the primary tool for this.

PROC MEANS DATA=dataset_name [options];
   VAR variable(s);
   CLASS variable(s);
   OUTPUT OUT=output_dataset_name MEAN=new_variable_name;
RUN;

Understanding each component is key to using this powerful procedure to calculate an average using SAS.

Breakdown of the PROC MEANS Syntax

Statement	Meaning	Unit (Input Type)	Typical Range (Example)
PROC MEANS DATA=	Initiates the procedure and specifies the input dataset.	SAS Dataset Name	work.sales, sashelp.cars
VAR	Specifies the numeric variable(s) for which to calculate the average.	Variable Name	Revenue, Age, Test_Score
CLASS	Specifies categorical variables used to group the analysis. SAS will calculate a separate average for each level of the class variable(s).	Variable Name	Region, Gender, Product_Category
OUTPUT OUT=	Creates a new SAS dataset containing the results instead of printing them.	SAS Dataset Name	work.summary_stats
MEAN=	Used within the OUTPUT statement to name the new variable that will hold the calculated average.	New Variable Name	Avg_Revenue, Mean_Score

Practical Examples

Example 1: Simple Average Calculation

Let’s calculate the average `MPG_City` and `MPG_Highway` for all cars in the `sashelp.cars` dataset.

• Inputs: Dataset = `sashelp.cars`, Analysis Variables = `MPG_City MPG_Highway`
• Units: The inputs are variable names. The results will be in Miles Per Gallon.
• Generated SAS Code:

  PROC MEANS DATA=sashelp.cars MEAN MAXDEC=2;
     VAR MPG_City MPG_Highway;
  RUN;

• Result: SAS will produce a table showing the mean city and highway MPG for all 428 cars in the dataset, rounded to two decimal places.

Example 2: Average by Group and Output to Dataset

Here, we will calculate the average `Invoice` price for cars, grouped by `Origin` (Asia, Europe, USA), and save the results into a new dataset called `work.avg_invoice`.

• Inputs: Dataset = `sashelp.cars`, Analysis Variable = `Invoice`, Class Variable = `Origin`, Output Dataset = `work.avg_invoice`
• Units: Input variables are names. The result (`Avg_Invoice`) will be in currency units.
• Generated SAS Code:

  PROC MEANS DATA=sashelp.cars NOPRINT;
     CLASS Origin;
     VAR Invoice;
     OUTPUT OUT=work.avg_invoice MEAN=Avg_Invoice;
  RUN;
  
  PROC PRINT DATA=work.avg_invoice;
  RUN;

• Result: This code first calculates the average invoice for each origin without printing the results. It then stores these three averages in a new dataset and finally, `PROC PRINT` is used to display the contents of the newly created results dataset. For more details on exporting, see our guide on SAS Data Export Techniques.

How to Use This SAS Average Code Generator

This tool simplifies the process of writing `PROC MEANS` code. Follow these steps:

1. Enter Dataset Name: Type the library and member name of the SAS dataset you want to analyze (e.g., `work.mydata`).
2. Specify Analysis Variable(s): Input the name(s) of the numeric columns you want to average. For multiple variables, just separate them with a space.
3. (Optional) Add Classification Variables: If you need to calculate the average for different groups, enter the categorical variable(s) here. For example, using `Region` would calculate a separate average for each region in your data.
4. (Optional) Name an Output Dataset: If you want to save your results to a new dataset for further analysis, provide a name here. Leaving this blank will print results to the standard output window.
5. Copy and Use: The complete, ready-to-run SAS code will appear in the results box. Click “Copy Code” and paste it into your SAS program editor.

Key Factors That Affect Average Calculations

• Missing Values: By default, SAS’s `PROC MEANS` and `AVG()` function ignore missing values (represented as a `.`) in the calculation. This is crucial because it prevents missing data from incorrectly skewing the average towards zero.
• Data Type: The average can only be calculated for numeric variables. Running `PROC MEANS` on a character variable will result in an error in the SAS log.
• Grouping with CLASS vs. BY: Using the `CLASS` statement is generally more efficient and straightforward for grouped analysis. A `BY` statement can also be used, but it requires the data to be sorted first using `PROC SORT`. Our guide on Advanced SAS Procedures covers this in depth.
• Weighting: If certain observations should contribute more to the average than others (e.g., in survey data), a `WEIGHT` statement can be used in `PROC MEANS` to specify a weighting variable.
• Choice of Procedure: While `PROC MEANS` is standard, `PROC SQL` provides an alternative way to calculate averages using the `AVG()` function with a `GROUP BY` clause, which can be more intuitive for those with a SQL background.
• Subsetting Data: Using a `WHERE` statement or `WHERE=` dataset option will filter the data *before* the average is calculated, which can significantly change the result compared to calculating an overall average.

Frequently Asked Questions (FAQ)

How do I calculate the average for all numeric variables in a dataset?

Simply omit the `VAR` statement. `PROC MEANS DATA=mydataset; RUN;` will calculate summary statistics for every numeric variable.

What’s the difference between MEAN and AVG?

In `PROC MEANS`, `MEAN` is the keyword for the statistic. In `PROC SQL` or a `DATA` step, `AVG()` is the function name. They both calculate the arithmetic mean.

How can I format the resulting average to two decimal places?

In `PROC MEANS`, add the `MAXDEC=2` option to the main statement: `PROC MEANS DATA=sashelp.cars MAXDEC=2;`. In an output dataset, you can apply a format using a `FORMAT` statement.

Why is my result blank or showing a dot (.)?

This happens if all values for the variable (or for a specific group) are missing. SAS correctly calculates the average of nothing as a missing value.

Can I calculate other statistics at the same time?

Yes. Simply add other statistic keywords to the `PROC MEANS` statement, such as `STD` (standard deviation), `MIN` (minimum), `MAX` (maximum), and `N` (count). For example: `PROC MEANS DATA=… MEAN STD MIN MAX;`.

How do I handle multiple grouping variables?

Just list them in the `CLASS` statement: `CLASS Region Department;`. SAS will calculate averages for every unique combination of region and department.

What does the NOPRINT option do?

The `NOPRINT` option tells SAS not to display the results in the output window. It’s most commonly used when you only care about creating an output dataset with the `OUTPUT` statement.

Is PROC SUMMARY the same as PROC MEANS?

They are very similar. The main difference is that `PROC MEANS` prints results by default, while `PROC SUMMARY` only creates an output dataset by default (it acts as if `NOPRINT` is on). Their syntax is otherwise nearly identical.