Base SAS 9.1 Procedures Guide, Volumes 1, 2, 3 and 4

Laying Out a Report

Planning the Layout

Report writing is simplified if you approach it with a clear understanding of what you want the report to look like. The most important thing to determine is the layout of the report. To design the layout, ask yourself the following kinds of questions:

When you understand the layout of the report, use the COLUMN and DEFINE statements in PROC REPORT to construct the layout.

The COLUMN statement lists the items that appear in the columns of the report, describes the arrangement of the columns, and defines headers that span multiple columns. A report item can be

Omit the COLUMN statement if you want to include all variables in the input data set in the same order as they occur in the data set.

Note: If you start PROC REPORT in the windowing environment without the COLUMN statement, then the initial report includes only as many variables as will fit on one page.

The DEFINE statement (or, in the windowing environment, the DEFINITION window) defines the characteristics of an item in the report. These characteristics include how PROC REPORT uses the item in the report, the text of the column header, and the format to use to display values.

Usage of Variables in a Report

Much of a report s layout is determined by the usages that you specify for variables in the DEFINE statements or DEFINITION windows . For data set variables, these usages are

A report can contain variables that are not in the input data set. These variables must have a usage of COMPUTED.

Display Variables

A report that contains one or more display variables has a row for every observation in the input data set. Display variables do not affect the order of the rows in the report. If no order variables appear to the left of a display variable, then the order of the rows in the report reflects the order of the observations in the data set. By default, PROC REPORT treats all character variables as display variables.

Order Variables

A report that contains one or more order variables has a row for every observation in the input data set. If no display variable appears to the left of an order variable, then PROC REPORT orders the detail rows according to the ascending , formatted values of the order variable. You can change the default order with ORDER= and DESCENDING in the DEFINE statement or with the DEFINITION window.

If the report contains multiple order variables, then PROC REPORT establishes the order of the detail rows by sorting these variables from left to right in the report. PROC REPORT does not repeat the value of an order variable from one row to the next if the value does not change, unless an order variable to its left changes values.

Across Variables

PROC REPORT creates a column for each value of an across variable. PROC REPORT orders the columns by the ascending, formatted values of the across variable. You can change the default order with ORDER= and DESCENDING in the DEFINE statement or with the DEFINITION window. If no other variable helps define the column (see COLUMN Statement on page 908), then PROC REPORT displays the N statistic (the number of observations in the input data set that belong to that cell of the report).

If you are familiar with procedures that use class variables, then you will see that across variables are class variables that are used in the column dimension.

Group Variables

If a report contains one or more group variables, then PROC REPORT tries to consolidate into one row all observations from the data set that have a unique combination of formatted values for all group variables.

When PROC REPORT creates groups, it orders the detail rows by the ascending, formatted values of the group variable. You can change the default order with ORDER= and DESCENDING in the DEFINE statement or with the DEFINITION window.

If the report contains multiple group variables, then the REPORT procedure establishes the order of the detail rows by sorting these variables from left to right in the report. PROC REPORT does not repeat the values of a group variable from one row to the next if the value does not change, unless a group variable to its left changes values.

If you are familiar with procedures that use class variables, then you will see that group variables are class variables that are used in the row dimension.

Note: You cannot always create groups. PROC REPORT cannot consolidate observations into groups if the report contains any order variables or any display variables that do not have one or more statistics associated with them (see COLUMN Statement on page 908). In the windowing environment, if PROC REPORT cannot immediately create groups, then the procedure changes all display and order variables to group variables so that it can create the group variable that you requested . In the nonwindowing environment, it returns to the SAS log a message that explains why it could not create groups. Instead, it creates a detail report that displays group variables the same way as it displays order variables. Even when PROC REPORT creates a detail report, the variables that you define as group variables retain that usage in their definitions.

Analysis Variables

An analysis variable is a numeric variable that is used to calculate a statistic for all the observations represented by a cell of the report. (Across variables, in combination with group variables or order variables, determine which observations a cell represents.) You associate a statistic with an analysis variable in the variable s definition or in the COLUMN statement. By default, PROC REPORT uses numeric variables as analysis variables that are used to calculate the Sum statistic.

The value of an analysis variable depends on where it appears in the report:

Note: Be careful when you use SAS dates in reports that contain summary lines. SAS dates are numeric variables. Unless you explicitly define dates as some other kind of variable, PROC REPORT summarizes them.

Computed Variables

Computed variables are variables that you define for the report. They are not in the input data set, and PROC REPORT does not add them to the input data set. However, computed variables are included in an output data set if you create one.

In the windowing environment, you add a computed variable to a report from the COMPUTED VAR window.

In the nonwindowing environment, you add a computed variable by

Interactions of Position and Usage

The position and usage of each variable in the report determine the report s structure and content. PROC REPORT orders the detail rows of the report according to the values of order and group variables, considered from left to right in the report. Similarly, PROC REPORT orders columns for an across variable from left to right, according to the values of the variable.

Several items can collectively define the contents of a column in a report. For instance, in Figure 43.7 on page 872, the values that appear in the third and fourth columns are collectively determined by Sales, an analysis variable, and by Department, an across variable. You create this kind of report with the COLUMN statement or, in the windowing environment, by placing report items above or below each other. This is called stacking items in the report because each item generates a header, and the headers are stacked one above the other.

Sales Figures for Perishables in Northern Sectors ______Department_______ Sector Manager Meat/Dairy Produce Perishable Total -------------------------------------------------------- Northeast Alomar 0.00 .00 6.00 Andrews 0.00 5.00 5.00 Northwest Brown 0.00 .00 3.00 Pelfrey 5.00 .00 1.00 Reveiz 0.00 .00 0.00

Figure 43.7: Stacking Department and Sales

When you use multiple items to define the contents of a column, at most one of the following can be in a column:

More than one of these items in a column creates a conflict for PROC REPORT about which values to display.

Table 43.1 on page 872 shows which report items can share a column.

Note: You cannot stack order variables with other report items.

Table 43.1: Report Items That Can Share Columns
 

Display

Analysis

Order

Group

Computed

Across

Statistic

Display

 

X [*]

X

Analysis

 

X

X

Order

 

Group

 

X

 

Computed variable

 

X

 

Across

X [*]

X

 

X

X

X

Statistic

X

X

 

X

 

[*] When a display variable and an across variable share a column, the report must also contain another variable that is not in the same column.

When a column is defined by stacked report items, PROC REPORT formats the values in the column by using the format that is specified for the lowest report item in the stack that does not have an ACROSS usage.

The following items can stand alone in a column:

Note: The values in a column that is occupied only by an across variable are frequency counts.

Statistics That Are Available in PROC REPORT

Descriptive statistic keywords

CSS

PCTSUM

CV

RANGE

MAX

STDDEVSTD

MEAN

STDERR

MIN

SUM

N

SUMWGT

NMISS

USS

PCTN

VAR

Quantile statistic keywords

MEDIANP50

Q3P75

P1

P90

P5

P95

P10

P99

Q1P25

QRANGE

Hypothesis testing keyword

PROBT

T

These statistics, the formulas that are used to calculate them, and their data requirements are discussed in Keywords and Formulas on page 1354.

To compute standard error and the Student s t -test you must use the default value of VARDEF=, which is DF.

Every statistic except N must be associated with a variable. You associate a statistic with a variable either by placing the statistic above or below a numeric display variable or by specifying the statistic as a usage option in the DEFINE statement or in the DEFINITION window for an analysis variable.

You can place N anywhere because it is the number of observations in the input data set that contribute to the value in a cell of the report. The value of N does not depend on a particular variable.

Note: If you use the MISSING option in the PROC REPORT statement, then N includes observations with missing group, order, or across variables.

Using Compute Blocks

What Is a Compute Block?

A compute block is one or more programming statements that appear either between a COMPUTE and an ENDCOMP statement or in a COMPUTE window. PROC REPORT executes these statements as it builds the report. A compute block can be associated with a report item (a data set variable, a statistic, or a computed variable) or with a location (at the top or bottom of the report; before or after a set of observations). You create a compute block with the COMPUTE window or with the COMPUTE statement. One form of the COMPUTE statement associates the compute block with a report item. Another form associates the compute block with a location in the report (see Using Break Lines on page 876).

Note: When you use the COMPUTE statement, you do not have to use a corresponding BREAK or RBREAK statement. (See Example 2 on page 965, which uses COMPUTE AFTER but does not use the RBREAK statement). Use these statements only when you want to implement one or more BREAK statement or RBREAK statement options (see Example 9 on page 986, which uses both COMPUTE AFTER MANAGER and BREAK AFTER MANAGER.

The Purpose of Compute Blocks

A compute block that is associated with a report item can

A compute block that is associated with a location can write a customized summary.

In addition, all compute blocks can use most SAS language elements to perform calculations (see The Contents of Compute Blocks on page 874). A PROC REPORT step can contain multiple compute blocks, but they cannot be nested.

The Contents of Compute Blocks

In the windowing environment, a compute block is in a COMPUTE window. In the nonwindowing environment, a compute block begins with a COMPUTE statement and ends with an ENDCOMP statement. Within a compute block, you can use these SAS language elements:

ARRAY

IF-THEN/ELSE

assignment

LENGTH

CALL

RETURN

DO (all forms)

SELECT

END

sum

For information about SAS language elements see the appropriate section in SAS Language Reference: Dictionary .

Within a compute block, you can also use these PROC REPORT features:

Four Ways to Reference Report Items in a Compute Block

A compute block can reference any report item that forms a column in the report (whether or not the column is visible). You reference report items in a compute block in one of four ways:

Note: Referencing variables that have missing values leads to missing values. If a compute block references a variable that has a missing value, then PROC REPORT displays that variable as a blank (for character variables) or as a period (for numeric variables).

The following table shows how to use each type of reference in a compute block.

If the variable that you reference is this type

Then refer to it by

For example

group

name [*]

Department

order

name [*]

Department

computed

name [*]

Department

display

name [*]

Department

display sharing a column with a statistic

a compound name [*]

Sales.sum

analysis

a compound name [*]

Sales.mean

any type sharing a column with an across variable

column number [**]

_c3_

[*] If the variable has an alias, then you must reference it with the alias.

[**] Even if the variable has an alias, you must reference it by column number.

Compute Block Processing

PROC REPORT processes compute blocks in two different ways.

Note: PROC REPORT recalculates computed variables at breaks. For details on compute block processing see How PROC REPORT Builds a Report on page 950.

Using Break Lines

What Are Break Lines?

Break lines are lines of text (including blanks) that appear at particular locations, called breaks , in a report. A report can contain multiple breaks. Generally , break lines are used to visually separate parts of a report, to summarize information, or both. They can occur

Break lines can contain

Creating Break Lines

There are two ways to create break lines. The first way is simpler. It produces a default summary. The second way is more flexible. It produces a customized summary and provides a way to slightly modify a default summary. Default summaries and customized summaries can appear at the same location in a report.

Default summaries are produced with the BREAK statement, the RBREAK statement, or the BREAK window. You can use default summaries to visually separate parts of the report, to summarize information for numeric variables, or both. Options provide some control over the appearance of the break lines, but if you choose to summarize numeric variables, then you have no control over the content and the placement of the summary information. (A break line that summarizes information is a summary line.)

Customized summaries are produced in a compute block. You can control both the appearance and content of a customized summary, but you must write the code to do so.

Order of Break Lines

You control the order of the lines in a customized summary. However, PROC REPORT controls the order of lines in a default summary and the placement of a customized summary relative to a default summary. When a default summary contains multiple break lines, the order in which the break lines appear is

  1. overlining or double overlining (in traditional SAS monospace output only)

  2. summary line

  3. underlining or double underlining (in traditional SAS monospace output only)

  4. blank line

  5. page break.

In traditional SAS monospace output only, if you define a customized summary for the same location, then customized break lines appear after underlining or double underlining.

The Automatic Variable _BREAK_

PROC REPORT automatically creates a variable called _BREAK_. This variable contains

Using Compound Names

When you use a statistic in a report, you generally refer to it in compute blocks by a compound name like Sales.sum. However, in different parts of the report, that same name has different meanings. Consider the report in Output 43.1. The statements that create the output follow. The user -defined formats that are used are created by a PROC FORMAT step on page 963.

libname proclib ' SAS-data-library '; options nodate pageno=1 linesize=64 pagesize=60 fmtsearch=(proclib); proc report data=grocery nowindows; column sector manager sales; define sector / group format=$sctrfmt.; define sales / analysis sum format=dollar9.2; define manager / group format=$mgrfmt.; break after sector / summarize skip ol; rbreak after / summarize dol dul; compute after; sector='Total:'; endcomp; run;

Output 43.1: Three Different Meanings of Sales.sum

The SAS System 1 Sector Manager Sales Northeast Alomar 6.00 [1] Andrews ,045.00 --------- --------- Northeast ,831.00 [2] Northwest Brown 8.00 Pelfrey 6.00 Reveiz ,110.00 --------- --------- Northwest ,454.00 Southeast Jones 0.00 Smith 0.00 --------- --------- Southeast 0.00 Southwest Adams 5.00 Taylor 3.00 --------- --------- Southwest ,048.00 ========= ========= Total: ,313.00 [3] ========= =========

 

Here Sales.sum has three different meanings:

[1]  

In detail rows, the value is the sales for one manager s store in a sector of the city. For example, the first detail row of the report shows that the sales for the store that Alomar manages were $786.00.

[2]  

In the group summary lines, the value is the sales for all the stores in one sector. For example, the first group summary line shows that sales for the Northeast sector were $1,831.00.

[3]  

In the report summary line, the value $6,313.00 is the sales for all stores in the city.

Note: Unless you use the NOALIAS option in the PROC REPORT statement, when you refer in a compute block to a statistic that has an alias, you do not use a compound name. Generally, you must use the alias. However, if the statistic shares a column with an across variable, then you must reference it by column number (see Four Ways to Reference Report Items in a Compute Block on page 875).

Using Style Elements in PROC REPORT

Using the STYLE= Option

Table 43.3: Locations and Default Style Elements for Each Statement in PROC REPORT

Statement

Valid Location Values

Default Location Value

Default Style Element

PROC REPORT

REPORT, COLUMN, HEADERHDR, SUMMARY, LINES, CALLDEF

REPORT

Table

BREAK

SUMMARY, LINES

SUMMARY

DataEmphasis

CALL DEFINE

CALLDEF

CALLDEF

Data

COMPUTE

LINES

LINES

NoteContent

DEFINE

COLUMN, HEADERHDR

COLUMN and HEADER

COLUMN: Data

HEADER: Header

RBREAK

SUMMARY, LINES

SUMMARY

DataEmphasis

style-element-name

style-attribute-specification(s)

Specifications in a statement other than the PROC REPORT statement override the same specification in the PROC REPORT statement. However, any style attributes that you specify in the PROC REPORT statement and do not override in another statement are inherited. For instance, if you specify a blue background and a white foreground for all column headings in the PROC REPORT statement, and you specify a gray background for the column headings of a variable in the DEFINE statement, then the background for that particular column heading is gray, and the foreground is white (as specified in the PROC REPORT statement).

Using a Format to Assign a Style Attribute Value

You can use a format to assign a style attribute value. For example, the following code assigns a red background color to cells in the Profit column for which the value is negative, and a green background color where the values are positive:

proc format; value proffmt low-<0='red' 0-high='green'; run; ods html body=' external-HTML-file '; proc report data=profits nowd; title 'Profits for Individual Stores'; column Store Profit; define Store / display 'Store'; define Profit / display 'Profit' style=[background=proffmt.]; run; ods html close;

Controlling the Spacing between Rows

Users frequently need to shrink a report to fit more rows on a page. Shrinking a report involves changing both the font size and the spacing between the rows. In order to give maximum flexibility to the user, ODS uses the font size that is specified for the REPORT location to calculate the spacing between the rows. Therefore, to shrink a table, change the font size for both the REPORT location and the COLUMN location. Here is an example:

proc report nowindows data= libref.data---set-name style(report)=[font_size=8pt] style(column)=[font=(Arial, 8pt)];

Printing a Report

Printing with ODS

Printing reports with the Output Delivery System is much simpler and provides more attractive output than the older methods of printing that are documented here. For best results, use an output destination such as Printer or RTF. For details on these destinations and on using the ODS statement, see The Complete Guide to the SAS Output Delivery System .

Printing from the REPORT Window

By default, if you print from the REPORT window, then the report is routed directly to your printer. If you want, you can specify a form to use for printing (see Printing with a Form on page 883). Forms specify things like the type of printer that you are using, text format, and page orientation.

Note: Forms are available only when you run SAS from a windowing environment.

Operating Environment Information: Printing is implemented differently in different operating environments. For information related to printing, consult SAS Language Reference: Concepts . Additional information may be available in the SAS documentation for your operating environment.

Printing with a Form

To print with a form from the REPORT window:

  1. Specify a form. You can specify a form with the FORMNAME command or, in some cases, through the File menu.

  2. Specify a print file if you want the output to go to a file instead of directly to the printer. You can specify a print file with the PRTFILE command or, in some cases, through the File menu.

  3. Issue the PRINT or PRINT PAGE command from the command line or from the File menu.

  4. If you specified a print file, then do the following:

    1. Free the print file. You can free a file with the FREE command or, in some cases, through Print utilities in the File menu. You cannot view or print the file until you free it.

    2. Use operating environment commands to send the file to the printer.

Printing from the Output Window

If you are running PROC REPORT with the NOWINDOWS option, then the default destination for the output is the Output window. Use the commands in the File menu to print the report.

Printing from Noninteractive or Batch Mode

If you use noninteractive or batch mode, then SAS writes the output either to the display or to external files, depending on the operating environment and on the SAS options that you use. Refer to the SAS documentation for your operating environment for information about how these files are named and where they are stored.

You can print the output file directly or use PROC PRINTTO to redirect the output to another file. In either case, no form is used, but carriage control characters are written if the destination is a print file.

Use operating environment commands to send the file to the printer.

Printing from Interactive Line Mode

If you use interactive line mode, then by default the output and log are displayed on the screen immediately following the programming statements. Use PROC PRINTTO to redirect the output to an external file. Then use operating environment commands to send the file to the printer.

Using PROC PRINTTO

PROC PRINTTO defines destinations for the SAS output and the SAS log (see Chapter 36, The PRINTTO Procedure, on page 787).

PROC PRINTTO does not use a form, but it does write carriage control characters if you are writing to a print file.

Note: You need two PROC PRINTTO steps. The first PROC PRINTTO step precedes the PROC REPORT step. It redirects the output to a file. The second PROC PRINTTO step follows the PROC REPORT step. It reestablishes the default destination and frees the output file. You cannot print the file until PROC PRINTTO frees it.

Storing and Reusing a Report Definition

The OUTREPT= option in the PROC REPORT statement stores a report definition in the specified catalog entry. If you are working in the nonwindowing environment, then the definition is based on the PROC REPORT step that you submit. If you are in the windowing environment, then the definition is based on the report that is in the REPORT window when you end the procedure. SAS assigns an entry type of REPT to the entry.

In the windowing environment, you can save the definition of the current report by selecting File – Save Report A report definition may differ from the SAS program that creates the report (see the discussion of OUTREPT= on page 894).

You can use a report definition to create an identically structured report for any SAS data set that contains variables with the same names as the ones that are used in the report definition. Use the REPORT= option in the PROC REPORT statement to load a report definition when you start PROC REPORT. In the windowing environment, load a report definition from the LOAD REPORT window by selecting File – Open Report

Related

Категории