@MEDIAN

Returns the median (the middle number) of the specified data set. Half the numbers in the data set are larger than the median, and half are smaller.

Syntax

@MEDIAN (SKIPNONE | SKIPMISSING | SKIPZERO | SKIPBOTH, XrangeList)
Parameter Description

SKIPNONE

Includes all cells specified in the data set, regardless of their content, during calculation of the median.

SKIPMISSING

Excludes all #MISSING values from the data set during calculation of the median.

SKIPZERO

Excludes all zero (0) values from the data set during calculation of the median.

SKIPBOTH

Excludes all zero (0) values and #MISSING values from the data set during calculation of the median.

XrangeList

A list of numeric values across which the median is calculated. Referred to generically throughout this topic as “the data set.

Can be a valid member name, a comma-delimited list of member names, cross dimensional members, or a return value from a member set function or range function (including @XRANGE).

For more information about XrangeList, see Range List Parameters in the topic Range and Financial Functions.

Notes

  • If the member you are calculating and the data set (XrangeList) are not in the same dimension, use the @RANGE or @XRANGE function to cross the member with the list of members (for example, to cross Sales with the children of 100).

  • @MEDIAN sorts the data set in ascending order before calculating the median.

  • When the data set contains an even number of values, the @MEDIAN function calculates the average of the two middle numbers.

  • @MEDIAN treats #MISSING values as 0 unless SKIPMISSING or SKIPBOTH is specified.

  • When you use @MEDIAN in a calculation script, use it within a FIX statement. Although FIX is not required, using it may improve calculation performance.

  • When you use @MEDIAN across a large range in a sparse dimension, you may need to increase the size of the calculator cache. For more information on the calculator cache, see the Oracle Essbase Database Administrator's Guide.

Example

The following example is based on the Sample Basic database. Assume that the Measures dimension contains an additional member, Median. This example calculates the median sales values for all products and uses the @RANGE function to generate the data set:

FIX (Product)
Median = @MEDIAN(SKIPBOTH,@RANGE(Sales,@CHILDREN(Product)));
ENDFIX

This example produces the following report:

                         Jan     New York
                      Actual       Budget
                      ======       ======
Sales    Colas         678          640
         Root Beer     551          530
         Cream Soda    663          510
         Fruit Soda    587          620
         Diet Drinks   #MI          #MI
          Product     2479         2300

Median   Product       625          575

Because SKIPBOTH is specified in the calculation script, the #MI values for Diet Drinks are skipped. The remaining four products create an even-numbered data set. So, to calculate Median->Product->Actual, the two middle numbers in the set (587 and 663) are averaged to create the median (625). To calculate Median->Product->Budget, the two middle numbers in the set (530 and 620) are averaged to create the median (575).

The following example assumes a Year dimension is added to Sample Basic. It calculates median using cross-dimensional members in the data set.

FIX(Product)
Median = @MEDIAN(@XRANGE("2011"->"Sep", "2012"->"Mar"));
ENDFIX

The above calculation is performed across the following multidimensional range specified by XrangeList:

2011->Sep
2011->Oct
2011->Nov
2011->Dec
2012->Jan
2012->Feb
2012->Mar

See Also