Catálogo de publicaciones - libros

You may recall that we touched briefly upon member properties back in Chapter 1. Now we’re going to take a further look at them and see how the value of these properties can be maximized using MDX.

Many times in business situations we want to know exactly how many somethings we have; the somethings may be customers, for instance, or products. It’s a straightforward enough question to a human being and you might expect the answer to be simply found too, but in the past it has proved surprisingly troublesome. All that’s behind us now as MDX has a splendidly useful Distinct Count function.

In Chapter 14 we will be looking at a variety of features that are best demonstrated using a Parent-Child dimension. The good news is that these dimensions are easy to understand, the bad is that Chapter 14 makes no sense at all unless you are familiar with them. Of course, in your OLAP travels you may already have become intimately acquainted with this type of dimension, in which case please feel free to skip this and dive immediately into Chapter 14.

The topics covered in this and the next chapter all come under the general heading of issues which typically need to be addressed in a financial application, so that is how we’ve illustrated them. However, once again, we want to stress that this is simply a convenient way of illustrating them — the topics can be applied in a host of different applications.

We’re still using the Budget cube and in the last chapter we successfully dealt with three problems using Custom Order, Custom Rollup and Custom Members. However, this cube continues to have some problems and, by a remarkable stroke of good fortune, resolving those problems enables us to introduce you to some further MDX-related topics.

An action, according to the MDX function list in Analysis Services’ help system, is an operation that can be initiated by an end user upon a selected cube or portion of a cube. Right, fine — but what does it mean? This is one occasion where a ‘for instance’ is worth its weight in help system entries, so here we go.

Color coding is a means of guiding the users’ investigation of a cube by using colors to bring attention to certain values. You may have come across something similar in Excel: in a stock control spreadsheet for instance, as soon as a value for stock holding drops below a certain level, that value can be displayed in red, making it much easier to spot and reducing the likelihood of running out of stock completely. Color coding in Analysis Services is broadly similar — making it easier to spot patterns, trends or a value that has drifted from the norm.

In Chapter 3 we used queries to introduce the general syntax used in MDX. As we said in that chapter, most of the time people use some form of GUI tool to query the cube, so they won’t be writing MDX queries by hand. Nevertheless, we have included this chapter on more advanced querying for three reasons.

The first edition of this book (18 chapters long) was written as an easy introduction to MDX. Assuming we succeeded in this task, by now you should be over that initial steep learning curve. So, when we came to add this extra material for the second edition we felt that it was appropriate to change gear a little and look at some more advanced topics. Recursion and NonEmptyCrossJoin came pretty much top of the list of advanced topics that people find challenging, so we decided to cover those.

The term ‘recursion’ is used to describe something that references itself. My favorite definition of recursion is found in the joke.