Definition of process instance-independent measures

Process instance-independent measures are defined in the client-specific measure configuration (XML file with the document type definition keyindicatorconfiguration.dtd) in the general context of data series.

Definition of data series

A data series (pikicube XML element) consists of process instance-independent measures and referenced dimensions. It must contain at least one process instance-independent measure (pikidef) and at least one referenced dimension (refdim). Referenced dimensions must be dimensions configured in the PPM system.

Process instance-independent measures in data series are always of the Process type in order to ensure maximum usability in the PPM system. Therefore, the type of the data series itself is not important, see chapter Usage (type) of a data series.

For each data series, at least one referenced dimension must be marked as a key dimension (refdim ... iskeydimension="TRUE"). By default, all referenced dimensions are key dimensions. A particular value combination of the specified key dimension(s) supplies a unique data row within a data series, i.e., a particular value combination exists only once within a data series. 

Example

The following data series contains three data rows that differ by the value combinations of the specified key dimensions (*):

D_COUNTRY*

D_PLANT*

D_DEPARTMENT*

D_RECORDED BY

SALES

COSTS

Germany

Hamburg

42

Smith

400000

 

Germany

Frankfurt

17

Hartmann

510000

360000

USA

Pittsburgh

53

Fox

 

410000

Each of the three data rows can occur only once within the data series. The specific value combination of the key dimensions D_COUNTRY, D_PLANT, and D_DEPARTMENT (e.g., Germany; Hamburg; 42) represents the identifier of a data row.

Configuration

The following general file structure illustrates the configuration of a data series (pikicube XML element):

<?xml version="1.0" encoding="ISO-8859-1"?>

<!DOCTYPE keyindicatorconfig SYSTEM

"KeyindicatorConfiguration.dtd">

<keyindicatorconfig>

...

<pikicube name="...">

<description language="de" name="..."/>

...

<pikidef name="..." retrievertype="..." Line break

dimreferring="...">

<description language="de" name="...">

Descriptive text (tooltip)

</description>

...

<datatype name="..."/>

</pikidef>

<refdim name="..." refinement="..." Line break

iskeydimension="TRUE"/>

...

</pikicube>

...

</keyindicatorconfig>

The following tables explain the configuration of a process instance-independent data series:

ELEMENT and
ATTLIST pikicube

Description

pikicube

Process instance-independent data series

name

Data series name unique in the system. It is also used as the name of the cube in the database.

comment
(optional)

Comment on the data series; used in PPM Customizing Toolkit.

editable

editable="TRUE" (default value) enables data input for the data series in the Configuration/Data input module of the PPM interface.

type

Usage (type of data series) that determines which dimensions may be used in the data series as referenced dimensions. The default value is PROCESS, i.e., only process dimensions (dimtype="PROCESS" in the dimension definition) may be specified in the data series.

Other valid values:

OT_FUNC (only function dimensions allowed in the data series)

RELATION (only dimension of the RELATION type allowed in the data series)

For more information, please refer to chapter Usage (type) of a data series.

relname

Only for type="RELATION". A single relation existing in the PPM system is to be specified with its name, e.g. relname="REL_WORKS_TOGETHER". The data series is assigned to the specified relation.

deletedata Line break
onredefinition

Obsolete,
no longer used.

description

Language-specific description of the data series. The description must be specified in at least the default language.

pikidef

Definition of a process instance-independent measure, at least one for each data series, see below. 

refki

Obsolete, no longer used.

refdim

Referenced dimension, see below 

ELEMENT and
ATTLIST refdim

Description

refdim

A dimension existing in the PPM system, to which the process instance-independent measures of the data series refer. You must specify at least one referenced dimension for each data series.

For process instance-independent measures internal dimensions are not supported as referenced dimensions (refdim).

name

Internal name of the dimension existing in the PPM system.

refinement

Dimension step width that data import is to be performed with.

The dimension values to be imported must be specified in this step width exactly.

iskeydimension

iskeydimension="TRUE" (default value) specifies that the referenced dimension is a key dimension of the data series. The value combinations of all specified key dimensions render each data row of a data series unique.

ELEMENT and
ATTLIST pikidef

Description

pikidef

Definition of a process instance-independent measure. You must specify at least one definition for each data series.
A process instance-independent measure can be used in a single data series.

name

Name of the process instance-independent measure that is unique in the system.

type

Obsolete, no longer used.

retrievertype

Measure retriever type. Default value: KEYINDICATOR (averaging)

Further values:
NUM_KEYINDICATOR (summation)
FREQ_KEYINDICATOR (obsolete, is no longer used)
FACTORY_KEYINDICATOR (is no longer used)

dimreferring

Type of dimension reference
LOOSE: Loose
STRICT: Strict
Default value: LOOSE

kigroup
(optional)

Assignment of the process instance-independent measure to an existing measure group

description

Language-specific description of a process instance-independent measure. The description must be specified in at least the default language.

datatype

Data type of the values of a process instance-independent measure

You can define any number of data series (pikicube XML elements) within a measure configuration. In a data series, you can specify any number of process instance-independent measure definitions (pikidef XML elements). The same dimension reference applies to all process instance-independent measure definitions of a data series (refdim XML elements).

You can conveniently configure process instance-independent data series in the sub-module Process instance-independent measures in the CTK module Measures and dimensions.

You can export all process instance-independent data series configured in a PPM system via the XML interface using the runppmconfig command line program with the parameter -keyindicator to an XML file.

Example 1

<?xml version="1.0" encoding="ISO-8859-1"?>

<!DOCTYPE keyindicatorconfig SYSTEM

"KeyindicatorConfiguration.dtd">

<keyindicatorconfig>

...

<pikicube name="PIKICUBE_TURNOVER_PROD_GROUP">

<pikidef name="TURNOVER_PROD_GROUP" Line break

retrievertype="KEYINDICATOR" Line break

dimreferring="STRICT" Line break

kigroup="KI_GROUP_COST">

<description language="de"

name="Umsatz pro Produktgruppe"/>

<description language="en"

name="Turnover by product group"/>

<datatype name="DOUBLE"/>

</pikidef>

<refdim name="TIME" refinement="BY_MONTH"/>

<refdim name="D_PRODUCT_GROUP"/>

<refdim name="PROCESSTYPE" refinement="BY_LEVEL2"/>

</pikicube>

...

</keyindicatorconfig>

A data series with the internal name PIKICUBE_TURNOVER_PROD_GROUP is created.

The definition (pikidef) of the process instance-independent measure TURNOVER_PROD_GROUP specifies a strict dimension reference (dimreferring="STRICT") and an assignment of the process instance-independent measure to the KI_GROUP_COST measure group.

The process instance-independent measure (refdim="...") strictly refers to the TIME, D_PRODUCT_GROUP, and PROCESSTYPE dimensions. The reference to the PROCESSTYPE dimension is defined with the detailed (refinement="BY_LEVEL2") step width.
Since the iskeydimension attribute is not specified in the refdim elements, the default attribute value TRUE is used, i.e., all referenced dimensions are used as key dimensions of the data series. 

Example 2

...

<pikicube name="PIKICUBE_COSTS">

<description language="en" name="Costs"/>

<pikidef name="OVERHEAD_COSTS" Line break

retrievertype="KEYINDICATOR" Line break

dimreferring="LOOSE">

<description language="en" name="Overhead costs"/>

<datatype name="COST"/>

</pikidef>

<refdim name="PROCESSTYPE" refinement="BY_LEVEL2" Line break

iskeydimension="FALSE"/>

<refdim name="TIME" refinement="BY_MONTH" Line break

iskeydimension="TRUE"/>

<refdim name="MATERIAL" refinement="BY_LEVEL2" Line break

iskeydimension="TRUE"/>

</pikicube>

...

This file extract defines the data series Costs with the process instance-independent measure Overhead costs of the data type COST with the internal name OVERHEAD_COSTS which is unique in the PPM system.

The measure value retriever type is averaging (KEYINDICATOR) and the dimension reference is loose (LOOSE).
In the refdim XML elements, the PPM dimensions TIME and MATERIAL are specified as key dimensions (iskeydimension="TRUE") of the dimension reference for the process instance-independent data series.
Additionally, step widths that differ from the default step widths are specified for the dimension values to be imported.

Registration of process instance-independent measures at the PPM system

Process instance-independent measures are registered in the process tree (useki element in the XML file with the document type definition keyindicatorprocesstree.dtd) at process type groups and process types.

Further information on registering process instance-independent measures at the process tree is available in chapter Register measures and dimensions of process instance-independent data series.