Build real world planning, budgeting, and forecasting solutions with a collection of over 60 simple but incredibly effective recipes with this book and ebook. I received a copy of the IBM Cognos TM1 Cookbook from Packt Publishing and was asked to write a review on this Forum and also Amazon. IBM Cognos TM1. Cookbook. Build real world planning, budgeting, and forecasting solutions with a collection of over 60 simple but incredibly effective recipes.
|Language:||English, Spanish, Indonesian|
|Genre:||Business & Career|
|ePub File Size:||17.59 MB|
|PDF File Size:||12.86 MB|
|Distribution:||Free* [*Regsitration Required]|
IBM Cognos TM1 Developer's Certification Guide. Read more IBM Cognos Business Intelligence Dashboarding Cookbook. Read more. IBM Cognos TM1 Cookbook Build real world planning, budgeting, and forecasting solutions with a collection of over 60 si. Chapter 3. Building Cubes and ViewsIn this chapter, we will cover:Re-orderin.
There are three types of pick list: Please note that all the aggregations are performed at run time in RAM. We have also defined subsets which can be reused while designing other TM1 objects. This cube-based technology does fast number crunching at the server end and helps end customers move away from manual and tedious planning, budgeting, and analytics processes. Setting security and access rights. Once editing is done, chores should be reactivated and can be executed at the scheduled date and time.
Moving data between versions. Moving data using scripts and parameters. Creating and scheduling a chore. Customizing Drill Paths. Creating a drill-through path to a view in another cube. Creating a drill path to an ODBC source. Using Rules for Advanced Modelling. Creating the Expenses cube. Examining a Spread Profile cube, Lookup cube, and Virtual cube. Moving balances from last month to next month. Converting Currencies.
Creating currency dimension and attributes. Creating a currency cube. Populating currency attributes using control cubes.
Creating rules to apply exchange rates and calculate an equivalent amount in target currency. Modelling for Different Fiscal Requirements.
Adding a year dimension. Creating a continuous time model. Optimizing Rules Performance. Troubleshooting rules and feeders.
Working with Managed Planning Applications. Installing Apache Tomcat 6. Contributing data to managed planning application. Navigating the TM1 Contributor workspace. Opening a node and modifying a view. Entering, editing, and distributing data. Reviewing data. Creating and deploying managed planning applications to the Web.
Setting security and access rights. Assigning security to managed planning applications. Defining Workflow. Adding users and groups to the TM1 Security. Reviewing submitted tasks. You've finished your project on Click here to start other projects, or click on the Next Section link below to explore the rest of this title. Buy eBook Buy from Store. Font size rem 1. Share Facebook Email Twitter Reddit. Doing custom calculations in dimensional hierarchies. Automating Data Load Processes using Scripts.
Loading data from multiple columns into a single column. Creating a dimension and a subset in Turbo Integrator. Moving data between versions. Moving data using scripts and parameters. Creating and scheduling a chore. Customizing Drill Paths. Creating a drill-through path to a view in another cube. Creating a drill path to an ODBC source.
Using Rules for Advanced Modelling. Creating the Expenses cube. Examining a Spread Profile cube, Lookup cube, and Virtual cube. Moving balances from last month to next month. Converting Currencies. Creating currency dimension and attributes. Creating a currency cube.
Populating currency attributes using control cubes. Creating rules to apply exchange rates and calculate an equivalent amount in target currency. Modelling for Different Fiscal Requirements. Adding a year dimension. Creating a continuous time model. Optimizing Rules Performance. Troubleshooting rules and feeders.
Working with Managed Planning Applications. Installing Apache Tomcat 6. Contributing data to managed planning application. Navigating the TM1 Contributor workspace. Opening a node and modifying a view. Entering, editing, and distributing data. Reviewing data. Creating and deploying managed planning applications to the Web. Hence profit booked can change for different regions, time period and products.
Again time dimension can be structured in a year, month, and day format, which form a hierarchy for time dimension. Year, month, and day can be different levels of the time dimension. Other dimensions can also be structured into multiple levels; however, it's not necessary for a dimension to have more than one level.
Similarly, total revenue can be another measure or a measure dimension, both imply the same thing in TM1 terminology.
It is interesting to note that in TM1 a measure dimension may not be always different from a regular analysis dimension. Hence, we can have both numeric and string values in a measure dimension.
The only difference lies in what needs to be viewed and what acts as perspective to view that. For example, product type can be a measure dimension in one case where we want to know the most popular product for different region and time period combinations. Otherwise, it would be a regular dimension when we want to measure total booked revenue for region and time period combinations.
Such structure once defined leads to creation of cubes and hence analysis of our measures booked profit from different perspectives dimensions like time, region, product types. Imagine an analysis needs to be performed for total revenue year time for Asia Pacific region Region.
Now that we know the figure, say million USD, we want to further drill down to see how the figure is distributed among different products. Since time dimension has different levels year as the first level , we can also know the monthly breakup of these figures, and to do that we just need to drill down from year level to month level Jan' , Feb' , Dec' Drilling up from month level to year level should also be possible in such a case say from Jan' to In a nut shell, depending on the organizations needs and availability of data, dimensions and measures are defined.
Dimensions are then structured in different levels to form a hierarchy. Once such structure is in place multi-dimensional model is ready. A cube can be created that is the same as organizational data, but in the multi-dimensional format, the structure of which is defined by the model. Cubes are then shared among the users, who can analyze data as per business needs. Therefore, in the preceding example products, time and region are dimensions regular dimension.
Out of these time is a multi-level dimension and structured in a hierarchy, where days are rolled up into months and months are rolled up in a year year month day. Total revenue and booked profit are measures measure dimension.
Elements are actual data which comprises the dimensions and hence cubes. For example, time dimension is made up of elements like , January, 21 December , which are again structured in a hierarchy as just explained. Product dimension is made up of elements such as vehicles, FMCG, and so on. Total revenue and total booked profit are measure dimensions and are elements in themselves. As just explained, elements are used to define data in the cubes.
These act as coordinates that identify a certain numeric value or a text string in a cube. Elements are basic building blocks of cubes and define what data is actually stored in a cube.
Every element has a type associated with it in TM1 as shown in the following table: It doesn't have any child elements. Sales Revenue, 21 Dec assuming time dimension has day as the last level String Elements which define those cells that have string data.
One or more children elements roll up to a consolidated element. Hence, weight defines contribution of an element in an aggregation.
For example, if we consider a consolidated element, Total Sale Price, for a product which we intend to calculate as summation of Base Price and Tax, then each simple element that is Base Price and Tax contributes equally with a factor 1.
Here 1 can be called as weights of respective simple elements Base Price and Tax. Similarly different simple elements can contribute differently to aggregated elements. Negative weights are also possible for a case that requires subtraction, for example: Hence, here Marked Price has weight as 1 while Discount has -1 as its weight. The same results can be achieved using rules, but using weights and consolidations in the preceding manner to achieve the output is much faster than rules.
Please note that all the aggregations are performed at run time in RAM. Cube only sores leaf level, simple, not null elements. As we just learned elements identify data in a cube.
Element attributes describe the elements themselves. Attributes are used to describe elements. These are used to describe features of an element, aliases or alternative names, format, and precision. We can select elements by attribute values in the subset editor, which we will see in later sections.
We can also display element names in the TM1 window using their aliases. To create attributes and assign attribute values, we use attributes editor and will be described in later sections. Alias Attributes, for example an element January may be called differently in German, French, and Chinese are different aliases of the same element. Display Format Attributes, for example a numeric measure element may be formatted by applying different format attributes such as currency, precision, percentage, and so on.
Please note that attributes just contribute and provide supporting data for an element. It is recommended to apply display formats to measure dimension the measures we track in the cube using either attributes editor or cube viewer window. TM1 determines which display format to use in the cube viewer window, as follows: If dimensions are stacked, then TM1 checks from the bottom upward.
If dimensions are stacked, then TM1 checks from right to left. The elements are inspected from right to left. When we want to list multiple attribute values for a single element, additional elements or dimensions should be created, otherwise we might lose valuable detail.
For example, if a car is available in multiple colors then we should consider creating additional elements for each color in a separate dimension. That way we can determine car sales for a particular color. However, if a car is available in a single color, then color of the car say red, should be an attribute rather than an additional element in a separate dimension. Dimension can be created manually or using turbo integrator processes. TM1 does support spaces and some special characters as names.
Utmost care should be taken while defining names to the dimension as these define where exactly a data cell is located in the cube. Hence, renaming a dimension will change how data points are aligned in the cube.
If we need to rename a dimension, it is advisable that we destroy and recreate the cube. Business-specific names with proper naming conventions across the organization should be used while naming and defining dimensions and elements. Dimensions can be created: We can create dimensions either manually or by using rules.
Getting ready We will use the demodata TM1 server we created in the previous chapter. Ensure that TM1 Admin Server service is started. Double-click on the demodata icon on desktop to start the data server, created as part of Chapter 1. Open the Architect tool from the Start menu. Double-click on demodata. For username enter "admin" and for password, do not enter anything.
Click on the OK button to view expended demodata object. Right-click on Dimensions and then click on Create New Dimension. It will open Dimensions Editor window. The same can be done by clicking on Dimensions and then clicking on Create New Dimension from the Dimensions menu on the toolbar.
In the Dimension Editor right-click and choose Insert Element from the menu. The same functionality can be achieved by clicking on Edit menu and choosing Insert Element. Keep the rest to default and click on Add button. Other available options in Element Type drop-down list are worth noting in context to element types explained in the previous section.
Click on Display Properties Window as shown in the following screenshot: After clicking on the icon, you will see the next window. It shows properties such as the Name, Security owner, Type, Level, and more. Click on Save button.
Please note change in the title bar. Now we will be adding a consolidated calculated item Profit as shown in the following screenshot. On the same window right-click to open the menu and select Insert Element as shown.
Insert a consolidated element by name of Profit and select Consolidated from the Element Type drop down. Click on Add and then OK. It will add a new element Profit. Please note that it is a consolidated item hence displayed differently. Now click on OK to close the Dimension Editor without saving the latest addition of Profit element, which we have added just to give an idea towards adding consolidated and calculated elements.
We will use more such elements in later recipes. How it works This measure dimension can now be included in a cube and data can be populated for its elements. Calculated and consolidated elements are calculated as and when required and as defined in the measure dimension. See also We have already seen creating a measure dimension manually.
In the same chapter we will be Creating a Months dimension, which we will use as a time dimension in the following section. We will also be Creating dimensions using the turbo integrator.
A Months dimension, when included in the cube will give time perspective to the data populated in the cube. Time dimension is desirable in almost all the use cases, in a practical scenario, to view and analyze data from time aspect. Getting ready Please ensure that as explained above TM1 Admin server service is started and demodata TM1 server is running.
How to do it Keep the Server Explorer window for demodata server open as shown in the following screenshot: Right-click Dimensions and click on Create New Dimension. Open Excel and type Jan in A1 cell as shown in the following screenshot: Drag the cursor vertically till A12 until all 12 months appear as shown. Copy the 12 months from Excel and paste them in the grey pane of Dimensional Editor. Save the dimension with the name Months and click on OK. These act as pointers to map data in cubes and maps to a numeric data value and not a text.
Now we will create another dimension by the name of Versions dimension as shown in the following screenshot. Please keep the Server Explorer window from the preceding recipe open as shown in the following screenshot: Right-click on Dimensions and click on Create New Dimension.
Click on OK and save the dimension as Versions. It is just a grouping of measures and should not be confused with regular dimensions which have descriptive data to define where each data point is mapped in the cube. Each data point described above can be termed as a measure. There's more Elements in a regular dimension which maps string data instead of numeric measures are called as string elements.
An example of such a string element is "Comments". We will create string elements later in the book. A dimension having at least one string element is called as string dimension. As a general practice string dimensions are created last.
Measure dimensions are created after regular dimensions as an accepted practice. Hence, it is advisable to create measure dimension first followed by regular dimensions and string dimension at last. We will be adding these levels manually using Dimension Editor. From the Edit menu click on Insert Element and add the following with default options.
In the preceding screenshot please note that though the weights are 0 they will be 1 when children are created in the dimension. The resulting dimensional structure should look like as shown in the following screenshot: Release when icon appears to move Jan, Feb, and Mar as child elements to Q1. Its value will be aggregation of child values which are Jan, Feb, and Mar. Similar is the case for other consolidated elements.
Same results can be achieved by creating Total Year. Again right-click on Q1 and insert Jan, Feb, Mar as its children, and so on. Right-click on Q1 and click on Element Properties to verify that. Please note that Element Type is set to Consolidated. Other Element Types are Simple and String as explained in previous sections.
Similarly, click on Jan to view its element properties which will be Simple. As explained in the preceding steps, repeat to make the hierarchy as shown in the following screenshot. Click on Set Dimension Order as shown in the following screenshot. Save the resulting dimension and close the Dimension Editor. Each element is assigned a unique index in a dimension. This is used to order the elements in a dimension. When we change the order of elements in a dimension we need to explicitly tell TM1 to assign new index numbers to each element by clicking on Set Dimension Order.
Next time when the dimension is opened, the same order is maintained. Adding Next and Previous attributes to the Months dimension In this section we will be adding text attributes to the existing Months dimension using Attribute Editor. Then we will demonstrate adding an alias attribute. Getting ready Make sure that TM1 Admin server service is started. We will be using demodata TM1 server to demonstrate the recipe hence make sure that the TM1 server is also started.
We will start with the Server Explorer window as seen before.
Open Server Explorer, right-click on Months dimension and select Edit Element Attributes as shown in the following screenshot. From the Edit menu select Add New Attribute. In the New Attribute dialog box type Name of the new attribute as Next with default options. Similarly, add another Text attribute called Previous. Now enter the values for attribute values as shown in the following screenshot. Add another attribute with name Full Name and of type Alias.
By default TM1 assigns default values to Alias attribute. These values can be changed as shown in Attribute Editor. Click on OK and close. Hence, we have added text attributes Next, Previous and alias attribute Full Name for Months dimension.
Click on Unit Sales Price and drag till Margin without releasing the mouse button to select three cells as shown. Then click on Format button. Click on Currency option and then click on OK.
Click on OK to close. Hence, we have done formatting of measure elements using Format attribute. In the preceding section, we demonstrated functionality related to attributes and have used them as text attributes for a dimension.
We also used the alias attribute and applied formatting to elements using format attributes. See also So far we have created dimensions manually and added attributes. In the next section we will be Creating dimensions using the Turbo Integrator. Turbo Scripts generated in the module, and data from the variety of sources can be Extracted, Transformed and Loaded in TM1 cubes. Turbo Integrator's basic object is the Turbo Integrator process which can be created to: In the next section, we will create dimensions using the TI process.
Further details will follow in later sections. We will be using demodata TM1 server to demonstrate the recipe so make sure that the TM1 server is also started. Now we will create a new Products dimension using TI processes. Create Products. Open and empty Excel sheet and enter data as shown in the following screenshot.
Save the file as Products. Please note the path may differ according to the location where TM1 is installed. The path where the CSV is kept can be according to user preferences. The preceding path is just an example and may be changed according to the specific preferences.
Open Server Explorer and right-click on Processes as shown in the following screenshot. Select Create New Process. Browse to the Products. Delimiter Type should be selected as Delimited and Delimiter should be selected as Comma. Click on Preview to view first 10 rows in the data source. Click on OK to dismiss the following warning, which is only relevant when working on a remote server. Please note that this is just an informative warning stating that in case data source is on the remote computer then the text file must be named according to UNC naming conventions.
In the demo we are following, text file is placed on local server; hence, the warning does not apply to us. Example of such an UNC naming path is stated as follows: After we dismiss the warning by clicking on OK and follow the preceding steps Turbo Integrator dialog should look as shown in the following screenshot: Click on Variables tab and modify Variable Name and Contents as shown in the following screenshot: Hence, Contents column for this row should be set to "Element".
The column indicates how the incoming data will be used, transformed, and mapped in the TI process. Other values which this column can have are as shown in the following table, with a description.
Hence, for more clarification on these please refer to the following table.
Contents Description Ignore Ignore the contents of the column Element Column is converted to leaf level elements Consolidation Column is converted to consolidated elements, and parent to other elements Data Column has data values for the cube Attribute Column contains element attributes Other Values do not fall in any of the above categories.
It is used by custom variables or formulas String values from first column in Products. Values read from the variable will be directly mapped to the leaf level elements in the Products dimension. These will be mapped to consolidated elements. No formula is required as the string values read from the CSV are directly mapped to the elements of the Products dimension.
Click on Maps tab and then on Dimensions tab. This tab specifies mapping from source to target dimension for leaf elements only. We are creating this dimension for the first time; hence, we let TI take Create action rather than Update or Recreate. Target element to be created will be of Numeric type as selected in the Element Type drop down. Another option here for Element Type is String, which as explained in the previous section will map to string descriptions in the cube, for example, Comments element can have strings, which are stored in the cube.
Above setting is directly mapped to "N" or "S" symbols, visible in dimensional or subset editor, as already explained in previous sections. Hence Element Type for vProducts is set as Numeric as elements created will serve as coordinates to a numerical measure instead of some string description. The Element Order specifies the sort order by which elements will be arranged in the final dimension. Internally, elements in the dimension are sorted according to the indices assigned to each of the elements.
The setting will determine what indices are assigned to which elements and hence the sort orders in the resulting dimension. The Element Order can be set to Automatic or Manual: The former setting enables users to sort elements according to the Hierarchy sorted by upper levels first and then within a level , Level sorted within the same level , or by Name in alphabetical order.
The latter setting allows users to arrange elements manually in the dimensional editor itself. Click on By Input button. Resulting Dimensions tab should look as shown in the following screeshot: Click on Consolidations tab. This tab specifies mapping from source to target dimension for consolidated elements. For vProductTypes row, click on Child Variable column and select vProducts as shown in the following screenshot: Click on Automatic for Select one of the sort. Resulting screen should look as shown in the following screenshot: As these are consolidated elements, child variables are specified for Child Variable column as shown in the preceding screenshot.
Weight in the above figure signifies by how much the specified child element contributes while rolling up to parent consolidated element.
Click on the Advanced tab and then click on Prolog, Metadata, Data, and Epilog tabs so that underlying scripts are updated, reviewed, and compiled. See also We have just created dimensions manually and using Turbo Integrator processes. Cubes can now be created using the dimensions and arranging them in a specific order. We will see more about creating cubes in Chapter 3 when we discuss Creating cubes.
Execution of scripts for a TM1 process There are four scripts which are executed while a TI process is kicked off. Each of the tabs shown in the preceding screenshot displays the code executed with each script.
These scripts can also be modified manually if absolutely required. More details about the scripts will be shared in later chapters. We will be using demodata TM1 server to demonstrate the recipe; hence, make sure that the TM1 server is also started.
It is recommended to save the process immediately when a new empty process is created. Using the TI wizard a process cannot be executed until all fields and tabs are completed without syntax errors. So saving in advance is very useful in many cases. Click on run to execute the process. Click on OK button and close the Turbo Integrator dialog box. In the Server Explorer window note a new dimension called Products has been created.
Double-click on the Products dimension to view the elements and associated hierarchy. These are the same elements which are in the Products. We have already ordered these according to the hierarchy in TI process; otherwise it would have been possible to order them in the Dimensional Editor window. Similarly, we can create a dimension based on Channels. Please take a note of the sample CSV as shown in the following screenshot: Resulting dimension should look as shown in the following screenshot: The preceding created dimension will be used for further recipes.
In the preceding steps we have created the Products and Channels dimension through Turbo Integrator process. Elements are created directly from corresponding CSVs. Therefore, in a similar fashion we can create dimensions out of data spread across a variety of data sources using Turbo Integrator processes. See also Chapter 3 will focus on Creating cubes and we will see Loading data using Turbo Integrator processes in Chapter 4, which will enable us to load data in the cubes.
We also created and used attributes as an alias to the elements and to format measure dimension. In this chapter, we will extend the same concepts and use these to create TM1 cubes. Multi-dimensional cubes are a data structure which allows manipulation and analysis of data from multiple perspectives.
A TM1 cube stores data for reports, analysis, and staging data for further calculations. Source data can come from multiple sources and be arranged in a multi-dimensional format called cube. There can be multiple cubes based on a model. Each cube contains data and business logic clubbed together, and the structure is defined by the underlying model, which again is composed of multiple dimensions, measures rules, and calculations. Building Cubes and Views Generally as per semantics, a TM1 cube should have at least one measure dimension and two or more regular dimensions, which in turn are made up of one or more elements.
Though in practical scenarios there can be cubes without a measure dimension. Elements from each dimension identify a single tuple or data point inside a cube. It's like a three dimensional axis system x, y, z each coordinate tuple identifies a specific data point in space. Cube can be thought of as an extension of the same three axis system, but with n number of dimensions instead of three as in the preceding example.
Additionally, as best coding practice please ensure the dimension order keeping in mind the following mentioned rule: If string element is not defined in the last dimension, it is ignored by the cube. A sparse dimension will have relatively lesser percentage of filled data points then a dense dimension. Memory Used is a measure available in the properties pane for cubes and dimensions and can serve as a pointer to decide sparsity.
In this chapter, you will also learn about pick lists which are a list of valid values that appear when a user wants to select valid values for a cell in the cube. These values are predefined and offer valid values which a specific element or a cube cell can have. When an administrator defines a pick list for an element or a cell, a drop-down menu containing predefined values appear for data entry, into the specified cell, when user is browsing a cube in one of the clients.
The user, while entering values for such elements and cube cells, can only select from predefined values defined in the pick list, and that is the way to provide validation checks to ensure the user cannot enter any ad hoc wrong values in the cube. Data type and format for the pick list and element must match, for administrator to define the pick list on the element or cell.
Please note that string pick list, which can only be applied to string elements, must also reside in last dimension in the cube, otherwise they won't appear. The data entered via Turbo Integrator processes or using Spreading, after pick list is defined, will not be validated by pick lists. So it is possible to see values in the cells that do not match with the ones in the pick list.
For example, if we have gender as a dimension, then it makes sense to provide users with only two valid options while they enter data for gender which are male and female. Anything other than that cannot be entered and TM1 will throw error if a user tries to enter anything other than what is defined in the pick list. This appears in every cube containing that dimension.
There are three types of pick list: If the elements of the subset change, value in the pick list changes correspondingly. A subset pick list is defined using the syntax: If the elements of the dimension change, values in the pick list changes same as before correspondingly. It is defined using the syntax: For MS Excel a space needs to be defined for blank.
Re-ordering dimensions in a cube It's not always possible to be completely familiar with the business data; hence, sparsity of dimensions in a cube is not always predictable. Cube data and the distribution also changes with time; therefore, order of dimensions based on sparsity may shift from ideal to less than ideal.
This results in less than optimal performance. It is important to note that it is not advisable to reorder dimensions frequently as the reorder operation is a very expensive operation in terms of memory consumption.
To address the above issue TM1 includes a feature that lets the user optimize the order of dimensions in a cube, thereby consuming less memory and improving performance.
How to do it… 1. In the Tree pane of the Server Explorer, select the cube to be optimized. Click on Cube, Re-order Dimensions. The Cube Optimizer dialog box opens as shown in the following screenshot: Select a dimension in the New Order of Dimensions list box. Click the up or down arrows to change the order of the dimension in the cube.
Click on Test. Note the value next to the Percent Change label. If this value is negative, the new order of dimensions consumes less memory and is therefore more efficient. Repeat steps 3 through 5 until the most efficient ordering of dimensions is achieved. There's more… Views are created on top of existing cubes to focus on a specific layout of dimensions in the cube. These can be saved for future references and signify a useful snapshot of dimensions and data, which can be updated with the latest data on a click of Recalculate button, while keeping the underlying arrangements of dimensions and filter the same.
The functionality of Recalculate button is similar to refresh and used to load the view with the latest data every time the user clicks on that. Unlike Public views, Private views are only accessible to the user who has created the view. Public views are available to all users based on security applied. Generally, an Administrator controls the creation and access to such views. View is functionally similar to database views which are quite well-known.
These do not have data but just store definition and provide a pigeon hole view inside the cubes, very specific to business requirements. We will be creating a default view on a multi-dimensional cube in the next recipe, Creating a multi-dimensional cube.
Creating a multi-dimensional cube Now we will be creating a cube and using demodata TM1 server. Getting ready Ensure that TM1 Admin service is started. The demodata TM1 server must be running and ready for business. Open Server Explorer window. Right-click on Cubes and click on Create new cube as shown in the following screenshot: Select the entire available dimension in left pane; double-click to move all dimensions to Dimensions in new cube in right pane as shown in the following screenshot.
We have used underscores for names in the example, but that is not absolutely required as spaces are allowed in TM1. Users are free to use spaces in TM1 object names as required. Change order of the dimensions using the Up and Down arrow, as shown. Please note that the order is complied with the already stated ordering rule in the previous section. Click on the Create Cube button. At this point of time cube is an empty box without any data.
Nevertheless, the structure given by dimensions and elements is there. The cube viewer window opens up to display dimensions in the title, row, and column areas. Click on Months dimension once to open the Subset Editor as shown in the following screenshot: Click on All icon to expand the dimension to the lowest level.
Now, click on Filter by Level , select lowest level that is 0 and click on OK. From the Subset menu click on Save As, for name type Months and select options as shown in the following screenshot: Hence, a subset called Months has been saved which can be used in other cubes and views as time dimension.
Click on OK to close the Subset Editor. Here in the preceding step we have just created a public subset which is available to all users who have at least read access to the associated dimension.
Admin privilege to the parent dimension is needed to create a public subset. A private subset on the other hand is available only to the user who creates them. Click on Recalculate button to populate latest data.