Feeds

Hands on with MDX

Dimension and measures (yes, but can it handle cubic furlongs?)

Top 5 reasons to deploy VMware with Tegile

Yes, yes, but where’s the CODE?

OK, here we go. This MDX query finds the total of apple sales in London in 2007:

SELECT
{ [Products].[Apples] } ON COLUMNS,
{ [Location].[London] } ON ROWS,
FROM [FruitMDX]
WHERE [Time].[2007]

The syntax is undeniably reminiscent of SQL, using SELECT, FROM and WHERE and, like an SQL query it returns an answer. However, unlike SQL, the main use for MDX is not querying (useful as it can be for that). In Fast Track to MDX a book about using MDX in SQL server 2000, of the 21 chapters there are but two on queries. The rest cover MDX expressions.

An MDX expression is a partial statement, a small, highly adaptable and powerful code section. The power comes from the language’s ability to 'understand' multi-dimensional data, and this is a radical difference between querying relational databases using SQL and multi-dimensional databases using MDX. In relational theory (and indeed practice) rows of data are inherently unordered. The concepts of the 'next' row or the 'previous' row are meaningless, and standard SQL has no functions for locating rows by position. In a multi-dimensional world, position is vital and MDX is well equipped to manipulate data with regard to its position.

For instance, MDX has a function called CurrentMember which identifies the cell in which you are currently interested. It acts dynamically so as your focus moves, CurrentMember moves too.

The green cell in the cube above could be referred to as:

(Location.Chicago, Products.Lemons, Time.[April])

(MDX syntax insists that member names containing numbers, spaces, special characters and keywords be in square brackets).

Now that the cell has your attention it could also be referred to as:

(Location.CurrentMember, Products.CurrentMember, Time.CurrentMember)

The huge advantage of a function that understands about position in a cube, particularly with respect to time, is that it makes analytical queries very easy to construct.

Take an apparently simple question like "How many pears have we sold since the beginning of the year?" Anyone who has played the SQL will know how painful this is to code because SQL knows nothing about position in a table. (Yes, I know we can use cursors, which are invaluable, but they are not part of standard SQL). An MDX expression to do this analysis comprises merely six words and some bracketry. Wow.

The function YTD stands for 'Year To Date'. It requires a member and returns a set of members that includes the member you gave it and all previous members at that level within the same year. If you give it the member April, it returns a set containing the values for that month and for Jan, Feb and March.

We can use the CurrentMember function to identify the member to pass to the YTD function:

YTD(Time.CurrentMember)

We want to add up the values in the set returned by YTD, so:

Sum(YTD(Time.CurrentMember))

does that, and lastly we identify the measure to use:

Sum(YTD(Time.CurrentMember), Measures.[Units Sold])

That’s the completed expression. (Oh alright, it's seven words but only because the measure has a space in its name). The code above is an MDX expression that creates a calculated measure: it generates new values from existing values in the data set. Calculated measures can be used for many purposes, like working out growth as a percentage or a monetary value or sales averages for comparison over time. They’re almost infinitely flexible.

But the really clever thing about our calculated measure is that it works everywhere in the Time dimension. If you're at the month level, say September 2007, it sums the Units Sold values for January to September 2007 inclusive, and if you’re at Q2-2006, it'll total Q1 and Q2 values in 2006. Using the CurrentMember function lets the calculated measure work dynamically, returning a result wherever you are in multi-dimensional space. Clever or what? ®

Providing a secure and efficient Helpdesk

More from The Register

next story
Preview redux: Microsoft ships new Windows 10 build with 7,000 changes
Latest bleeding-edge bits borrow Action Center from Windows Phone
Google opens Inbox – email for people too thick to handle email
Print this article out and give it to someone tech-y if you get stuck
Microsoft promises Windows 10 will mean two-factor auth for all
Sneak peek at security features Redmond's baking into new OS
UNIX greybeards threaten Debian fork over systemd plan
'Veteran Unix Admins' fear desktop emphasis is betraying open source
Google+ goes TITSUP. But WHO knew? How long? Anyone ... Hello ...
Wobbly Gmail, Contacts, Calendar on the other hand ...
DEATH by PowerPoint: Microsoft warns of 0-day attack hidden in slides
Might put out patch in update, might chuck it out sooner
Redmond top man Satya Nadella: 'Microsoft LOVES Linux'
Open-source 'love' fairly runneth over at cloud event
prev story

Whitepapers

Cloud and hybrid-cloud data protection for VMware
Learn how quick and easy it is to configure backups and perform restores for VMware environments.
A strategic approach to identity relationship management
ForgeRock commissioned Forrester to evaluate companies’ IAM practices and requirements when it comes to customer-facing scenarios versus employee-facing ones.
High Performance for All
While HPC is not new, it has traditionally been seen as a specialist area – is it now geared up to meet more mainstream requirements?
Three 1TB solid state scorchers up for grabs
Big SSDs can be expensive but think big and think free because you could be the lucky winner of one of three 1TB Samsung SSD 840 EVO drives that we’re giving away worth over £300 apiece.
Security for virtualized datacentres
Legacy security solutions are inefficient due to the architectural differences between physical and virtual environments.