What is MDX?

What is MDX?
Multi-dimensional expression MDX (multidimen1_expression) is used to access Analysis
Languages developed by services multi-dimensional structure data cube. Powerful and can be used to define Analysis Services
Computing and security rules. Just as you must use SQL to access a relational database
The same language. To access the multi-dimensional analysis structure such as cube, you must use the MDX language. Mdx is similar in syntax to SQL, but it has the ability to process an online Processing (OLAP) three-dimensional, multidimensional
Better data features. For most common tasks, you can use MDX to obtain information in a single stereo. Mdx
Provides new script functions, attribute hierarchies, automatic coexistence functions, subcube operations, new data definitions and data operation statements, and new functions and operators.

Different types of simplified information are called dimensions. Multi-dimensional simplified data structure is called cube, and simplified information is called measurement ). Each dimension has one or more dimensions and measurements.

Basic syntax
For DB2 alphabrovCubeThe basic syntax of the body-executed MDX query is as follows:

Select {axisspecification} on columns,
{Axisspecification} on rows
From cubename
Where (sliceritems)
Axisspecification is a set of one or more tuples. You can enter tuples in the form of a list, or you can use the crossjoin function to "generate" tuples.
Cubename is the name of the defined alphabox cube.
Sliceritems is a tuple (usually a list of members separated by commas) that filters the query result set. If multiple slice members exist, each member must come from a different dimension and cannot reference the dimension in any axis specified in the query.

select
[store]. [store country]. [USA]. children on columns,
[product]. [All products]. [food]. children on rows
from [sales]
where ([measures]. [unit sales])
in this case, there are two axis rules, one of which is the vertical number of output and the other is the number of output rows. Each rule specifies a part of the dimension.