Building the Essbase Outline [ 110 ] Attribute dimensions are always dynamically calculated, which means they do not store the data. This is always a benet as they do not affect the size or performance of the database. Attribute dimensions must always be associated with a standard dimension. Earlier, we briey saw the differences between standard dimensions and attribute dimensions. Sometimes, your data has attributes that make it desirable to perform further analysis on it. In the case of the Esscar Motor Company, it may be desirable to track a certain vehicle model by its color. You may wish to compare how that model sells in markets based on population size. Essbase database attribute dimensions can help you with this type of analysis. One of the most notable features of an attribute dimension is the fact that the entire dimension is a dynamically calculated dimension. Previously, we told you about dynamically calculated members in a standard outline. In an attribute dimension, all members are dynamically calculated. What this means is that Essbase does not need to store the data for an entire dimension! All of the data is calculated on the y when requested and goes away when the user is nished. This, as previously mentioned, is a huge benet when considering CPU performance and storage options. Here are a few facts about attribute dimensions: • An attribute dimension is a dimension that must be associated with a standard dimension. • An attribute dimension can only be associated with one standard dimension, but a standard dimension can be associated with many attribute dimensions. • An attribute dimension can only be associated with a Sparse standard dimension. This is an important point. To create an attribute dimension and associate it to a standard dimension, you need to do the following: 1. Create a new dimension and tag it as an attribute dimension. 2. Next, create an association between this new attribute dimension and the standard dimension you wish to track attributes from. 3. Add the zero level members to the attribute dimension that correspond to the zero level members of the standard dimension. 4. Lastly, associate the members from the standard dimension to the zero level attribute dimension members. This material is copyright and is licensed for the sole use by Paul Corcorran on 5th July 2009 8601 ave. p #1, , lubbock, , 79423 Download at Boykma.Com Chapter 3 [ 111 ] As this is an advanced Essbase function, we have provided a high-level look at attribute dimensions and their set up and use. For complete details, please refer to the documentation which was provided when you purchased your version of software. The following screenshot shows the TOTAL VEHICLES dimension with the new COLORS attribute dimension added: User Dened Attributes (UDA) An Essbase UDA is a descriptive word or tag about an outline member. Similar to an alias, the main difference is that an alias may only be attached to one member. A UDA can be attached to many members. What the UDA offers is a way to simplify and make the ongoing operations of your database more efcient. This material is copyright and is licensed for the sole use by Paul Corcorran on 5th July 2009 8601 ave. p #1, , lubbock, , 79423 Download at Boykma.Com Building the Essbase Outline [ 112 ] For example, because forecasting analytics is one of the most common uses for an OLAP system, the users deal with time periods that fall into three distinct categories. In any Essbase database outline, you will have historical time periods, actual time periods, and forecast time periods. Typically, the data you have previously loaded for historical time periods, and the data you just loaded for actual time periods, does not need to be fully calculated since it is real data and does not need to be derived or changed. On the other hand, your forecast data still needs to be calculated whenever you make a change to a component piece of data. If you tagged your calendar periods in the outline with UDAs of Forecast for the future periods, LatestActual for the rst month back, and History for all historical periods in the outline. If you move the UDAs once a month to compensate for the new month in time, you could code a calculation script that would never need to be updated since you would write the script to calculate forecast periods by looking at the Forecast UDA. This material is copyright and is licensed for the sole use by Paul Corcorran on 5th July 2009 8601 ave. p #1, , lubbock, , 79423 Download at Boykma.Com Chapter 3 [ 113 ] Some UDA rules to keep in mind: • You can assign many UDAs to a single member, but you cannot assign the same UDA to a member more than once. • You can use the same UDA on many different members. • A UDA cannot be the same name or word as a member name or alias name. • UDAs cannot be used on Shared Members or members in an attribute dimension. • UDAs run with the dimension they were created in and can only be assigned to other members in that same dimension. • UDAs are member specic and only apply to the single member they are attached to. Parents, children, and so on, of a member assigned a UDA are not covered unless they also have the UDA assigned to them as well. UDAs are added to the database outline members in much the same way as adding an alias or updating any other outline member property. When adding a UDA to an outline member, Essbase will display a list of UDAs available for the dimension in which the member resides. If the desired UDA does not exist, you may simply add it the rst time. Dynamic Time Series Once again, Essbase gives us a feature so cool that it makes us wonder how we lived without it! The Dynamic Time Series function is one of those features. We can't think of anyone who wouldn't use this functionality. Likewise, we can't think of anyone who wouldn't need this functionality. What the Dynamic Time Series does for you is automatically calculate period-to-date data based on your Time dimension. Using our example database for the Esscar Motor Company, we see that it is probably a good idea to know our year-to-date vehicle production, by both calendar year and model year. The Dynamic Time Series function does this for you. You can easily calculate period-to-date data for any period as dened in your Time dimension. This material is copyright and is licensed for the sole use by Paul Corcorran on 5th July 2009 8601 ave. p #1, , lubbock, , 79423 Download at Boykma.Com Building the Essbase Outline [ 114 ] In other words, Essbase can automatically calculate a cumulative number based on how you have the Time dimension set up. If you want to track how many vehicles you have built from the beginning of the year through the current month, the Dynamic Time Series can do this for you. If you were to build 1,000 units in January and 1,000 units in February, and 1,000 units in March, you would expect to see 1,000 units as your net number for each month's data pulled. But, if you want to know the total units through each month, you would expect to see 1,000 for January, 2,000 for February, and 3,000 through March. This is what a period-to-date calculation will give you. The Dynamic Time Series member is actually created for you when you associate a predened Dynamic Time Series member with the appropriate outline Generation. For example, to use the year-to-date Dynamic Time Series functionality, you would associate a year member with the Generation 4 (month's level) Calendar Periods dimension. When you request the data at the Dynamic Time Series member, Essbase will automatically sum the Generation 4 level members within the given year and return the results to you. Essbase has provided the following predened Dynamic Time Series functions. As always, if you have a special need, there are plenty of calculation functions available to allow you to code even the most complex custom calculation: • H-T-D (History-to-date) • Y-T-D (Year-to-date) • S-T-D (Season-to-date) • P-T-D (Period-to-date) • Q-T-D (Quarter-to-date) • M-T-D (Month-to-date) • W-T-D (Week-to-date) • D-T-D (Day-to-date) Here is an important item to consider when coding Dynamic Time Series functionality. Your Dynamic Time Series member can be associated with any Generation in your time dimension, except the zero level Generation (you cannot have months adding upto a month-to-date value). The steps to add a Dynamic Time Series member to the Time dimension of our Esscar database outline are: 1. In EAS, open the Esscar outline in the outline editor 2. Click on Outline | Dynamic Time Series. This material is copyright and is licensed for the sole use by Paul Corcorran on 5th July 2009 8601 ave. p #1, , lubbock, , 79423 Download at Boykma.Com . sole use by Paul Corcorran on 5th July 20 09 8601 ave. p #1, , lubbock, , 794 23 Download at Boykma.Com Building the Essbase Outline [ 114 ] In other words, Essbase can automatically calculate a cumulative. licensed for the sole use by Paul Corcorran on 5th July 20 09 8601 ave. p #1, , lubbock, , 794 23 Download at Boykma.Com Building the Essbase Outline [ 112 ] For example, because forecasting analytics. the sole use by Paul Corcorran on 5th July 20 09 8601 ave. p #1, , lubbock, , 794 23 Download at Boykma.Com Chapter 3 [ 111 ] As this is an advanced Essbase function, we have provided a high-level