5 Filtering and Selecting Data for Analyses

This chapter explains how to construct filters, selection steps, groups, and calculated items and how to use them to specify the data that is displayed in analyses and dashboards. This chapter contains the following topics:

Saving Objects as Inline or Named

This section describes saving objects as inline or named. It contains the following topics:

What are Inline or Named Objects?

As you work with certain objects, you can create other objects that are saved with those objects. When you save one object with another, that object is saved "inline." You can save filters, groups, and calculated items inline. For example, you can create a group as part of an analysis. When you save the analysis, the group is saved "inline" or along with the analysis.

In addition to saving these objects inline, you can save them as individual objects with the subject area in the Oracle BI Presentation Catalog. When you save an object on its own, its becomes a "named" object. Named objects provide reusability, because you can create one object and use it multiple times with any analysis, dashboard (for filters), or dashboard page (for filters) that contains the columns that are specified in the named object. When the named object is updated and saved, those updates are immediately applied to all objects where the named object is used.

For example, after you save a group inline with an analysis, you can save the group as its own object in the catalog. You can then apply that named group from the Catalog pane to other analyses.

What is the Folder Structure for Named Objects?

Named filters, groups, and calculated items are generally saved to their related subject area folder. By saving the objects to a subject area folder, you ensure that they are available when you create an analysis for the same subject area.

Named objects saved in the /My Folders folder are available only to you. Objects saved in the /Shared Folders folder are available to other users who have permission to access the folders. If a subject area folder does not exist in your /My Folders folder or within the /Shared Folders folder, then the subject area folder is created automatically. When you save the object, the "Save As dialog" will display a default save path to /My Folders/Subject Area Contents/<subject area>. However, the dialog's Folders area continues to display all instances of the subject area folder in the catalog.

Saving Filters as Inline or Named

When you create an inline filter in the Analysis editor: Criteria Tab's "Filters pane", you can optionally save the inline filter as a named filter. Named filters can also be created at the analysis level or as a stand-alone object from the global header.

A named filter can filter all or some of the analyses that are embedded in a dashboard or analyses that are embedded on the same dashboard page.

Saving Groups and Calculated Items as Inline or Named

You can save groups and calculated item as an inline or named object:

  • When you create a group or calculated item while editing and saving a view or while working in the "Compound Layout", the group or calculated item is saved inline with the analysis.

  • When you work in the Selection Steps pane:

    • You can save a group or calculated item that is within a step as a named object in the catalog.

    • You can save a set of steps or the resulting members list for a column as a named object. You cannot save a set of steps as a group is one of the steps includes a calculated item.

See "Adding a Group to Another Analysis" for information on adding a saved group to an analysis.

To save a calculated item or group as a named object to the catalog:

  1. Display the "Selection Steps pane".

  2. Click the link for the calculated item or group, then click Save Calculated Item As or Save Group As to display the "Save As dialog".

  3. Complete the dialog to save the object to the catalog.

To save a set of steps as a group to the catalog:

  1. Display the "Selection Steps pane".

  2. Click the Save Selection Steps button to the far right of the column name.

  3. Complete the "Save Selection Steps dialog" to save the group as an object to the catalog.

What are Filters and Selection Steps?

You use both filters and selection steps to limit the results that are displayed when an analysis is run, so that the results answer a particular question. Together with the columns that you select for an analysis, filters and selection steps determine what the results will contain. Based on the filters and selection steps, only those results that match the criteria are shown. For example, depending on the industry in which you work, you can use filters and selection steps to learn who are the top ten performers, what are the dollar sales for a particular brand, which are the most profitable customers, and so on.

Note:

Another kind of filter, called a prompt, can apply to all items in a dashboard. Prompts can be used to complete selection steps and filters at runtime. For information, see Chapter 6, "Prompting in Dashboards and Analyses."

How Do Filters and Selection Steps Differ?

Filters and selection steps are applied on a column-level basis and provide two methods for limiting the data in an analysis. A filter is always applied to a column before any selection steps are applied. Steps are applied in their specified order. Filters and selection steps differ in various ways.

Filters

Filters can be applied directly to attribute columns and measure columns. Filters are applied before the query is aggregated and affect the query and thus the resulting values for measures. For example, suppose that you have a list of members in which the aggregate sums to 100. Over time, more members meet the filter criteria and are filtered in, which increases the aggregate sum to 200.

Selection Steps

Selection steps are applied after the query is aggregated and affect only the members displayed, not the resulting aggregate values. For example, suppose that you have a list of hierarchical members in which the aggregate sums to 100. If you remove one of the members using a selection step, then the aggregate sum remains at 100.

You can create selection steps for both attribute columns and hierarchical columns. Selection steps are per column and cannot cross columns. Because attribute columns do not have an aggregate member, the use of selection steps versus filters for attribute columns is not as distinctive as for hierarchical columns. While measure columns are displayed in the Selection Steps pane, you cannot create steps for them so steps do not affect them. Measures are used to create condition steps for attribute and hierarchical columns, such as Sales greater than $1 million.

Applying Filters to Attribute Columns to Affect Hierarchical Columns

You can use a filter on a related attribute column to affect the display of members in a hierarchical column. For example, suppose a hierarchical column contains the levels Year, Quarter, and Month. Suppose that a filter exists on the attribute column that corresponds to the Year hierarchy level. If you create a filter on Year to limit it to 2008 and 2009, then when the hierarchical column is displayed in a view, only those two years are visible. This functionality depends on the way that the logical columns have been defined in the business layer of the subject area in the Oracle BI Administration Tool.

Working with Selections of Data

As you specify which data members to include in an analysis, you create selections of data from the data source. Each selection specifies the criteria for a set of members for a particular column, such as Product or Geography. Each selection consists of one or more steps. A step is an instruction that affects the selection, such as add Product members whose values contain the text "ABC." The order in which steps are performed affects the selection of data. Each step acts incrementally on the results from previous steps, rather than acting on all the members for that column.

You can view these selection steps in the "Selection Steps pane". Steps are created via the following means:

  • When you add a column to an analysis, a selection step is created automatically to start with all members, unless you explicitly add specific members. As you drag and drop column members in the Results tab to add to the analysis, steps are also created automatically. For example, suppose that you drag and drop the FY2007 and FY2008 members from the Year hierarchical column to a pivot table. The selection step "Add FY2007, FY2008" is created.

  • As you add groups and calculated items, steps are created automatically.

  • You can create steps directly using the Selection Steps pane, to refine the selection of data for a particular hierarchical column or attribute column. You can display the Selection Steps pane from various places including the Results tab, the Criteria tab, and some view editors.

Selection steps can be one of the following types:

  • Explicit list of members — A step can include a list of members for a column, such as Boston, New York, Kansas, South. For hierarchical columns, the members can be from different hierarchy levels. For attribute columns, the members can be from only that column.

  • Condition step — A step can specify that members are selected from a column based on a condition, which can be one of various types including based on measures or on top/bottom values. This member list is dynamic and determined at runtime.

  • Based on hierarchy step — A step for hierarchical columns that allows you to select the type of relationship with which to select members. You can select a family relationship (such as children of or parent of), a specific hierarchy level (for level-based hierarchies only), or a level relationship (for level-based hierarchies only).

  • Groups and calculated items — A step can include a group or calculated item. Groups and calculated items can be used only with Add steps; they cannot be used in Keep Only or Remove steps.

    For information, see "Working with Groups and Calculated Items".

Creating Selection Steps

You create steps in the "Selection Steps pane", which you can display in various places. The following procedure describes how to create steps in the Results tab.

To create selection steps:

  1. Display the "Analysis editor: Results tab".

  2. If the Selection Steps pane is not visible, then click the Show Selection Steps pane button in the toolbar to display it.

    The pane might also be collapsed at the bottom of the Results tab. Click the plus sign icon to expand it.

  3. For the column whose steps you want to define, click the Then, New Step link.

  4. From the menu, select the option for the step type that you want to create and complete the resulting dialog.

Editing Selection Steps

You can edit existing selection steps, as described in the following procedure.

To edit selection steps:

  1. Hover the mouse pointer over the step in the Selection Steps pane and click a button on the resulting toolbar.

    You can perform various tasks such as displaying a dialog for editing the step, deleting the step, or changing the order of the step in the list of steps.

    For a group or calculated item, click its name to display a menu with options for editing and saving.

Saving Selection Steps as a Group Object

If you have created a set of selection steps, then you can save and reuse the set as a group object, as described in "Saving Groups and Calculated Items as Inline or Named".

Working with Selection Steps and Double Columns

If your repository is configured for double columns, then you can create a selection step on a double column. To do so, select the display values for that column and the step is automatically evaluated using the code values that correspond to those display values.

If you use double columns, then use care with the "New Calculated Item dialog". You can include a positional operator in the custom formula for the calculated item, such as $1, which specifies the column from the first row in the data set. When you include a positional operator, the display values cannot be mapped to the code values when evaluating the formula.

For information on double columns, see "Understanding the Double Column Feature".

Creating or Editing Column Filters

Use the following procedure to create or edit a named or inline filter.

To create or edit a column filter:

Note:

If your repository is configured for double columns, and you want to use an operator other than is equal to / is in, is not equal to / is not in, or is between and specify code values rather than display values, then you should explicitly choose the code column rather than the display column.

For information on double columns, see "Understanding the Double Column Feature".

  1. To create a named filter, use the following sub-procedure:

    1. From the Oracle Business Intelligence Home page, locate the global header, hover over the New menu, and from the menu select Filter. The Select Subject Area dialog is displayed.

    2. From the Select Subject Area dialog, choose the subject area for which you want to create a filter. The "Filter editor" is displayed.

    3. From the "Subject Areas pane", double-click the column for which you want to create the filter. The "New Filter dialog" is displayed.

  2. To create an inline filter, use the following sub-procedure:

    1. Either create a new analysis or access an existing analysis for which you want to create a filter. Click the Criteria tab.

    2. Locate the "Filters pane" and from the Filters Pane's toolbar, click the Create a filter for the current subject area button. The analysis' selected columns are displayed in the cascading menu.

    3. Select a column name from the menu. Or select the More Columns option to access the"Select Column dialog" from which you can select any column from the subject area.

      After you selected a column, the "New Filter dialog" is displayed.

      Note:

      If you want to add a filter for a column located in a different subject area, then you must first add that subject area to the analysis by clicking the Add/Remove Subject Area button in the "Subject Areas pane".
  3. In the Operator field, choose an operator for the filter. The operator list from which you can choose is populated based on the type of column that you selected. For more information about each operator, including the is prompted and is based on the results of another analysis operator options, see "Operators".

  4. In the Value field, specify one or more values to use when applying the filter or condition. You can:

    • Type values, using a semi-colon to separate the values.

    • Select values from the list or calendar.

      To search for specific values, click Search in the list box. The "Select Values dialog" is displayed, where you can search for and select values.

    If your repository is configured for double columns, and you are creating the filter on a display column, then by default, you specify display values. However, if your organization allows the display of code values, then you can specify code values rather than display values, but only if you use one of the following operators:

    • is equal to / is in

    • is not equal to / is not in

    • is between

    To specify code values, select the Select by Code Column check box and then specify the values.

  5. If your repository is configured for double columns, and you are creating the filter on a display column and want to filter on display values rather than code values, then deselect the Filter by Code Column check box.

  6. Click the Add More Options button to add a SQL expression, Session variable, Repository variable, or Presentation variable to the filter. Note that if you are setting the filter's value with a SQL expression or variable, you should leave the Value field blank. For more information on variables, see "Using Variables".

  7. Select the Protect Filter check box to protect the filter's value from being overridden by a matching prompt's value or when the user navigates to another report within the analysis. When you navigate from one report to another report within an analysis, any prompt values that you specified in the first report can be transferred to the second report.

  8. To convert the filter to SQL statements, select the Convert this filter to SQL check box. The "Advanced SQL Filter dialog" is displayed.

    Note:

    This is a one-way conversion. After you select the Convert this filter to SQL box, you can no longer view and edit the filter in the Edit Filter dialog. After you convert the filter to SQL statements, you can only view and edit the filter item as SQL statements in the Advanced SQL Filter dialog.

    For more information about this option, see "Creating and Editing the SQL Statements for a Column Filter in an Analysis".

  9. When you are finished specifying the filter's criteria, click OK.

  10. Save the filter in one of the following ways:

    • From the Analysis Editor, you can select Save Analysis to save the filter as an inline filter.

    • From the Filter Editor, you can select Save Filter to save the filter as a named filter.

    For more information, see "Saving Objects as Inline or Named".

Working with the EVALUATE_PREDICATE Function

This section provides the following topics on working with the EVALUATE_PREDICATE function:

For more information about this function, see "EVALUATE_PREDICATE".

How Can I Use the EVALUATE_PREDICATE Function with a Filter?

You can add an EVALUATE_PREDICATE function as an inline filter clause. You cannot use this function with hierarchical columns. Use this function when you cannot create the filter clause that you want with the Oracle BI EE filter operators. This function is intended for database functions with a return type of Boolean, and can be used only for SQL functions. Support for EVALUATE_PREDICATE does not extend across all multidimensional data sources at this time. Also, you cannot use EVALUATE_PREDICATE with XML data sources.

Example

The following is an example of how you can use the EVALUATE_PREDICATE function. This example requests that Markets.Region values be greater than 14 characters. After it is run, this example returns any rows where the length of the data in the Region field is greater than 14 characters (for example, "SOUTHERN REGION").

SELECT
0,
"Paint"."Markets"."Region",
"Paint"."Markets"."District", 
"Paint"."Sales Measures"."Dollars"
FROM "Paint"
Where EVALUATE_PREDICATE('length(%1)>14',"Markets"."Region").
ORDER BY 1,2,3)

Adding the EVALUATE_PREDICATE Function to an Inline Filter

Use the following procedure to add the EVALUATE_PREDICATE function to an inline filter. Note that you cannot use this function with hierarchical columns. For more information see "How Can I Use the EVALUATE_PREDICATE Function with a Filter?" and "EVALUATE_PREDICATE".

To add the EVALUATE_PREDICATE function to an inline filter:

  1. Go to the Analysis Editor's"Filters pane" and click the More Options toolbar button, and select the Add EVALUATE_PREDICATE function option.

    The "New EVALUATE_PREDICATE Function dialog" is displayed.

  2. Enter the function's formula.

    For an example of entering correct syntax, see "How Can I Use the EVALUATE_PREDICATE Function with a Filter?"

  3. Click OK.

    The EVALUATE_PREDICATE function is displayed in the Filters pane.

Applying a Named Filter to an Analysis

Use the following procedure to apply a named filter to an analysis.

To apply a named filter to an analysis:

  1. Create or open the analysis to which you want to add a named filter.

  2. Within the Analysis Editor: Criteria tab, locate the "Catalog pane" and navigate to the appropriate folder that holds the named filter. Filters are usually saved within the Subject Area subfolder. For example, Shared Folders/Sales/Subject Area Contents/Paint/region_filter or My Folders/Subject Area Contents/Paint/my_region_filter.

  3. Select the named filter and click the Add More Options button. The "Apply Saved Filter dialog" is displayed.

  4. Specify how you want to add the named filter to the analysis. You can select one or both of the following options:

    • Select the Clear all existing filters before applying box to remove all existing filters from the analysis before adding the named filter.

    • Select the Apply contents of filter instead of a reference to the filter box to copy the actual contents of the filter to the analysis. Copying the contents enables you to manipulate the filter criteria without altering the saved filter. When you clear the Apply contents of filter instead of a reference to the filter box, a reference to the filter is added to the analysis. From the analysis, you can view but not alter the named filter's contents.

  5. Click OK.

    The filter is displayed in the Filters Pane.

Combining and Grouping Column Filters

Combining and grouping column filters enables you to create complex filters without requiring you to know SQL statements.

To combine a column filter with other column filters:

  1. Create or open a named filter or analysis that contains an inline filter.

  2. If you are working with a named filter, then locate the Saved Filter pane and confirm that the filter contains two or more filter items. If you are working with an inline filter, then locate the Filters Pane and confirm that the filter contains two or more filter items.

    After you add the second filter item, the AND operator is displayed before the second filter item.

    Note:

    If you want to group filters, then the filter must contain three or more filter items.
  3. To change an AND operator to an OR operator, click the word AND. You can toggle between AND and OR this way. Note the following information:

    • The AND operator means that the criteria specified in each filter must be met. This is the default method for combining column filters.

    • The OR operator means that the criteria that is specified in at least one of the column filters must be met.

  4. As you add filter items, click AND and OR operators as necessary to construct the appropriate filter combinations.

  5. Save the filter in one of the following ways:

    • From the Analysis Editor, you can select Save Analysis to save the filter as an inline filter.

    • From the Filter Editor, you can select Save Filter to save the filter as a named filter.

    For more information, see "Saving Objects as Inline or Named".

Using a Saved Analysis as a Filter

You can create a filter based on the values that are returned by another analysis. Any saved analysis that returns a column of values can be used to filter the matching column in an analysis.

To create a filter based on the results of another saved analysis:

  1. Create or open a named filter or analysis that contains an inline filter.

  2. If you are working with a named filter, then locate the Saved Filter pane and from the "Subject Areas pane", select the column for which you want to create a filter.

    If you are working with an inline filter, then locate the Filters Pane, and from the Filters Pane toolbar, click the Create a filter for the current Subject Area button and select the column for which you want to create the filter.

    The "New Filter dialog" is displayed.

  3. In the Operator field, select is based on the results of another analysis.

    The Saved Analysis, Relationship, and Use values in Column fields are displayed.

  4. In the Saved Analysis field, either enter the complete path to the saved analysis or click the Browse button to locate the analysis upon which to base the filter.

    The column names from the saved analysis are displayed in the Use Values in Column drop-down list.

  5. Select a column name from the Use Values in Column field, and in the Relationship field, select the appropriate relationship between the results and the column to be filtered.

  6. Click OK.

Creating and Editing the SQL Statements for a Column Filter in an Analysis

You can create and edit the logical SQL WHERE clause to be used as a filter. While generally not necessary, this feature is available for users who want advanced filtering capability. For descriptions of SQL clauses, see Appendix A, "Logical SQL Reference."

Note:

After you convert a filter to SQL statements, you can view and edit the filter item as SQL statements in the Advanced SQL Filter dialog, only. You can no longer view and edit the filter in the Edit Filter dialog.

To create and edit the SQL syntax of a column filter:

  1. Create or open a named filter or an analysis that contains an inline filter.

  2. If you are working with a named filter, then locate the Saved Filter pane and from the "Subject Areas pane", select the column for which you want to create a filter.

    If you are working with an inline filter, then locate the Filters Pane, and from the Filters Pane toolbar, click the Create a filter for the current Subject Area button and select the column for which you want to create the filter.

    The "New Filter dialog" is displayed.

  3. Specify the filter's criteria and select Convert this filter to SQL.

  4. Click OK.

    The "Advanced SQL Filter dialog" is displayed.

  5. Enter your modifications in the SQL field, and click OK. Not that after you convert the filter to a SQL statement, you can no longer view and edit the filter in the Edit Filter dialog.

Working with Groups and Calculated Items

You can create a group or calculated item as a way to display data in a table, pivot table, or graph. Groups and calculated items allow you to add new "members" to a column, when those members do not exist in the data source. These members are also known as "custom members."

  • A group is a static list of members that you select or a static or dynamic list that is generated by selection steps. A group is represented as a member. You can drill in a group that was created for a hierarchical column but not in one that was created for an attribute column.

    A group uses the existing aggregation function of the measure column with which it is displayed. The aggregation is performed up from the lowest detail level in the Oracle BI Server, but no values are double-counted.

  • A calculated item is a computation between members, which is represented as a single member that cannot be drilled. When you create a calculated item, you add a new member in which you have selected how it will be aggregated, such as Sum or Average or a custom formula. The calculation is performed at the aggregated level, not at the lowest detail level.

Both groups and calculated items become selection steps for the analysis. Therefore, they apply to all views for that analysis. Both groups and calculated items can be saved as inline or named objects. For information, see "Saving Objects as Inline or Named".

What are Groups?

A group is a user-defined member of a column. A group can be a list of members or a set of selection steps that can be executed to generate a list of members. All the members must be from the same attribute column or hierarchical column, and in a hierarchical column, members can be from different hierarchy levels. Groups are always displayed at the bottom of the column list in the order in which they were added (as shown in the Selection Steps pane). Groups can contain members or other groups.

You can save a group to the catalog and re-use it. For example, you can apply a group to analysis and dashboard column prompts or variable prompts. If a group is applied to a prompt, then the prompt presents either the group as a choice list option or members of the group as choice list options to the user at runtime. See "Overriding a Selection Step With a Column Prompt" for information on overriding selection steps with prompts when the user selects groups or column members.

What are Calculated Items?

A calculated item is a calculation between members, which is represented as a single member. A calculated item allows you to override the default aggregation rule that is specified in the Oracle BI repository, and for an existing analysis, the aggregation rule chosen by the designer. You can define a calculated item using a custom formula (which is the default) or by combining selected members with a function (for example, SUM).

A calculated item is a user-defined member of a column. A calculated item can contain members or other calculated items. Calculated items are always displayed at the bottom of the column list in the order in which they were added (as shown in the Selection Steps pane). You can save a calculated item to the catalog and re-use it.

A calculated item is useful when you want to view and manipulate a set of members as a single entity. For example, you might define calculated members for the following:

  • Key accounts in a geographic region

  • High-end products

  • Non-standard time periods, such as the first three weeks in September

You can create calculated items for columns in various places. Calculations differ from the default aggregation rules that are applied to measures, as described in "Adding Totals to Tables and Pivot Tables". Internally, calculated items are processed as SQL SELECT statements, and the indicated functions are performed on the result set. For more information about SQL functions, see Appendix A, "Logical SQL Reference.".

How Will Calculated Items Created in Previous Releases Be Upgraded?

In previous releases of Oracle BI EE, when you created a calculated item in a pivot table, that calculated item applied only to that pivot table for that analysis. If you upgrade from a previous release, then all calculated items are automatically converted to selection steps, which apply to all views for an analysis.

For more information, see "What are Filters and Selection Steps?"

Creating Groups and Calculated Items

Use the following procedure to create a calculated item or group.

To create a calculated item or group:

  1. Click the button to create a new calculated item or group.

    The button is available in multiple locations. The following list provides a few of these locations:

    • In the toolbar of the view editor, click the New Calculated Item button or the New Group button.

    • In the Layout pane, in the <view-type> Rows or Columns area, click the More Options button to the right of a column name, then select New Calculated Item.

    You can also create calculated items and groups by clicking the Then, New Step link in the "Selection Steps pane".

  2. Complete the appropriate dialog, either "New Calculated Item dialog" or "New Group dialog".

  3. When the calculated item or group is complete, click OK.

    If any errors are detected, then a message will be displayed. Correct the error and click OK again.

Editing Groups and Calculated Items

You can edit groups and calculated item in various ways:

  • In the "Selection Steps pane", click the link for the calculated item or group, then click Edit to display the appropriate dialog.

  • If you have saved the object to the catalog, then select the object in the Catalog pane and click Edit to display the appropriate dialog.

The value of a group or calculated item might be affected by filters and selection steps, as described in the following list:

  • Filters — A group or calculated item is evaluated using only those members that have not been removed using filters. For example, if you have a calculated item for SUM(EAST + WEST) but WEST is removed through a filter, then only the EAST sum is included for the calculated item. If all members have been removed, then a null value is returned, which shows as an empty cell in a table or pivot table.

  • Selection steps — When you create selection steps, you can add a group or a calculated item in a step. Subsequent Keep Only or Remove steps might reference members that were included in the group or calculated item.

    • A group list is affected by members that are kept or removed in subsequent steps, but the group outline value remains the same. For example, suppose the MyNewYork group contains Albany and Buffalo and its value is 100. Suppose Albany is removed in a later step. The value of the MyNewYork group remains at 100, but Albany is no longer listed with the group.

    • A calculated item is not affected by members that are kept or removed in subsequent steps, because removals can affect the components of the formula.

Adding a Group to Another Analysis

You can add a group (also known as a "saved selection" in other products) to the same column on which it was created in another analysis. The group can be either a list of members or a set of selection steps. For information on these groups, see "Saving Selection Steps as a Group Object".

The following procedure describes how to add the group using the Results tab, but you can do so anywhere that the Catalog pane is displayed. You can also add a group using the following means:

  • In the Edit Member Step dialog, select Start with Group or Calculated Item in the Action box, then select the group from the Catalog pane in the Available list.

  • In the Selection Steps pane, click Then, New Step for the appropriate column, select Add Groups or Calculated Items, select Select Existing Groups and Calculated Items, and select the group in the resulting dialog.

To add a group to another analysis via the Catalog pane:

  1. On the Results tab, display an analysis that contains the same column to which you want to apply the selections from a group.

  2. Locate the Catalog pane and navigate to the appropriate folder that holds the saved group.

  3. Select the group and click the Add More Options button in the toolbar of the Catalog pane.

  4. Select one of the following:

    • Add to add the group itself, which generates an outline value for the group. In the table or pivot table, you see the group name, which can be expanded to see its member values. (Default)

    • Add Members to add only the groups members to the analysis. You will not see the outline value.

    The group or its members are included as an "Add" step on the Selection Steps pane. You can reorder the steps as appropriate.

Nesting Groups and Calculated Items

As you work with groups and calculated items, you might want to "nest" them; that is, create a group within a group, for example. The following list provides various scenarios for nesting groups and calculated items:

  • Groups can be nested; that is, they can include other groups. Nested groups are "flattened." For example, suppose that the my_favorite_countries group includes the my_favorite_cities group. When you display and expand the my_favorite_countries group in a table, you will not see the my_favorite_cities group. Instead, you will see the member values of the my_favorite_cities group.

  • Calculated items can be nested; that is, they can include other calculated items.

  • Groups cannot contain calculated items nor can calculated items contain groups.

  • When you work with selection steps:

    • You can save selection steps that include groups or calculated items as either a series of steps whose results are generated at runtime or as results that are static and are redisplayed each time.

    • You can apply a group that is a saved selection of steps, using the Catalog pane. If you do so when the saved selection contains a step for a calculated item, then you can only add the members of the group. You cannot add the group itself.

Examples of Calculated Items

The examples and explanations in this section assume that you have a basic understanding of SQL statements and their syntax. The syntax shown in these examples applies to the Custom Formula function in the "New Calculated Item dialog". The examples are hypothetical. Not all possible calculated items are shown.

Example 5-1 shows the code required to obtain the value of the current measure, such as dollar sales, for each of the products SoftDrinkA, SoftDrinkB, and SoftDrinkC, and adds the values together.

This is equivalent to selecting Sum from the Function list, and then typing or clicking 'SoftDrinkA','SoftDrinkB','SoftDrinkC' to add them to the Function field.

Example 5-1 Obtaining the Value of the Current Measure

sum('SoftDrinkA','SoftDrinkB','SoftDrinkC')

Example 5-2 obtains the minimum current measure, such as dollars in sales, for SoftDrinkA or SoftDrinkB, whichever is lower.

In Example 5-1 and Example 5-2, each functional calculated item is performed for each member in the outer layer, such as the Product layer. For example, if Year and Product are positioned on an axis, and one of the preceding calculated items is built on the Product layer, then the results will be computed per year.

Example 5-2 Obtaining the Minimum Current Measure

min('SoftDrinkA','SoftDrinkB')

Example 5-3 obtains the values for each item in the outer layer, such as Year and Product, and adds them together.

Instead of specifying a named item for an attribute column, such as SoftDrinkA, you can specify $n or $-n, where n is an integer that indicates the item's row position. If you specify $n, then the measure is taken from the nth row. If you specify $-n, then the measure is taken from the nth to the last row.

For example, for dollar sales, $1 obtains the measure from the first row in the data set, and $-1 obtains the measure from the last row in the data set.

Example 5-3 Obtaining the Values of Each Item in the Outer Layer

sum(*) 

Example 5-4 obtains the current measure, such as dollar sales, of the item from the first, second, and third rows, and sums them.

Example 5-4 Obtaining the Values of Measures and Summing Them

sum($1,$2,$3)

Example 5-5 adds sales of SoftDrinkA, SoftDrinkB, and SoftDrinkC. Note that the two calculated items shown in the example are equivalent. That is, Sum is the default function; therefore, it can be omitted.

Example 5-5 Adding Sales Values

'SoftDrinkA' + 'SoftDrinkB' + 'SoftDrinkC'
sum('SoftDrinkA','SoftDrinkB','SoftDrinkC')

Example 5-6 adds sales of SoftDrinkA with sales of diet SoftDrinkA, then adds sales of SoftDrinkB with sales of diet SoftDrinkB, and returns the maximum of these two amounts.

Example 5-6 Adding Sales Values and Returning the Maximum

max('SoftDrinkA' + 'diet SoftDrinkA', 'SoftDrinkB' + 'diet SoftDrinkB')