This section describes how to create and use numeric expressions in DeepSee MDX.
In DeepSee MDX, a numeric expression can have any of the following forms:
A numeric literal. For example: 37
The literal cannot start with a decimal point; that is, you must include a leading 0 with any fractional values. For example, 0.1
is valid is valid, but .1
is not valid.
A percentage literal. For example: 10%
There must be no space between the number and the percent sign.
Many MDX functions return numeric values, including AVG
, and others. Also, the IIF
function can return numeric values; this function evaluates a condition and returns one of two values, depending on the condition.
An expression that uses mathematical operators to combine numeric expressions. For example: MEASURES.[%COUNT] / 100
DeepSee supports the standard mathematical operators: +
(division), and *
(multiplication). It also supports the standard unary operators: +
(positive) and -
You can use parentheses to control precedence.
In the expression, if any value is null, the expression evaluates to null.
If you divide a value by 0, DeepSee treats the result as null.
The MDX function IIF
can be useful in such expressions.
Note that the value of a member expression depends upon the measure that is currently in use. By default, this expression evaluates to the number of records that belong this member. In contrast, if a specific measure is in use, this expression evaluates to the aggregate value of that measure across those records.
The MDX identifier for a dimension, such as [gend]
Note that the value of this expression depends upon the measure that is currently in use. By default, this expression evaluates to the number of records in the cube. In contrast, if a specific measure is in use, this expression evaluates to the aggregate value of that measure across all records in the cube.
A reference to a pivot variable that contains a numeric value. To refer to a pivot variable, use the following syntax:
is the logical variable name. Do not enclose this expression with square brackets. This syntax is not case-sensitive; nor is the pivot variable name.
You can use numeric expressions in the following ways:
This section shows examples of some of the less common kinds of numeric expressions. The first example shows that a member expression has a numeric value:
SELECT gend.h1.gender.female ON 0 FROM patients
The next example is a variation of the preceding:
SELECT gend.h1.gender.female+100 ON 0 FROM patients
As noted earlier, the value of a member expression depends upon the measure that is in use:
SELECT gend.h1.gender.female ON 0 FROM patients WHERE MEASURES.[avg age]
SELECT gend.h1.gender.female+500 ON 0 FROM patients WHERE MEASURES.[avg age]
The next example shows the numeric value of a dimension:
SELECT allerd ON 0 FROM patients
As noted earlier, the value of such an expression depends upon the measure that is in use:
SELECT allerd ON 0 FROM patients WHERE MEASURES.[avg allergy count]
© 1997-2019 InterSystems Corporation, Cambridge, MA