Formula Extension (FOX) (SAP Library - BW Business Planning and Simulation)

Formula Extension (FOX)

Use

You use the SEM-BPS formula extension FOX within the formula function, in order to access transaction data of the planning package more flexibly than when you sequentially, uniformly process all data records with the help of available operators. In addition to the formula operators, the formula extension offers you the possibility to make statements dependent on conditions, to run through loops, and issue messages.

Formula extensions contain a few, easy to learn language elements, with which you can solve almost all planning tasks, which occur in practice. The syntax of this language element is very similar to the SAP programming language ABAP. Therefore, you can use any existing ABAP knowledge, without having to concern yourself with the complexity of a full programming language (for example, memory management, database access).

Integration

The formula extensions are not an independent SEM-BPS function type, they are an enhancement to the formulas. You can continue to create formulas in the usual way, without having to make use of the formula extensions.

Features

General

All formula extension statements must be completed with a period "." A statement can extend over more than one line in the formula editor, and several statements can be in the same line. No distinction is made between upper and lower case. The system automatically converts lower case entries into upper case entries.

Comments

You can create comments in the formula editor to explain the statements you programmed, and to make the code easier to read. To do this, set the indicator * in the first column of a row. This means that the row is ignored by the syntax check, and is skipped when executing the function.

Note

It is important that you write the indicator * in the first column of a row. Otherwise the system will interpret the indicator as a multiplication operator and report a syntax error.

Syntax

* comment text

Explanations

comment text:
Any character string, which is not interpreted by SEM-BPS as a statement. The character string cannot extend over several lines.

Data Declaration (Local Variables)

You can declare local variables for a parameter group in the formula editor. You use these variables for example to check the dynamic conditions at the runtime of the planning function, which are dependent on the current key figure values. First you could determine the largest or smallest posted value of a key figure within a loop using the transaction data. Then you can save this value in a variable and then generate new plan data independently of the variable value. The system initializes variables of the types F and I with the value 0.

Note

For the language elements FOREACH and MESSAGE it is absolutely necessary that reference to characteristics take place via the variables declared before. A direct reference to a characteristic using its name is not possible with these language elements.

Syntax

DATA varname TYPE typname.

Explanations

varname:

Name of the variable. The name can contain letters and figures, but must always begin with a letter. The name must differ from the reserved words of the formula extension. All language elements belong to this (for example DATA, IF) and all formula operators (for example SIN, CEIL, DECL).

typname:

Name of the data type for the data saved in the variables. The permitted types are divided into two classes: predefined types and context-specific types.

Predefined types:

I: Integer. Used to save whole number values. For example, you use this type for the declaration of control variables with which you can count the number of iterations in a loop construct.
F: Floating point. Used to save decimals with up to three decimal places; decimal point is the period ".". You use this type for all calculation operations.

Context-specific types:

Context-specific types are types, which refer to the key figures and characteristic in the planning level to which the formula function belongs. Which types are allowed depends on which characteristics of the planning level were included in the list of fields to be changed of the formula function. The correct type name is then identical with the technical name of a characteristic. A variable whose type refers to a characteristic of the planning level, can accept all values which are maintained for this characteristic.
The KEYFIGURE_NAME type is an exception: This type is only allowed, when you have included the generic entry "key figure name" in the list of fields to be changed. You can program a loop with a variable of this type, which iterates over all key figures of the planning level. In such a loop, you can for example assign all key figures the same value, or you can define conditions on which the value assignment should depend.

Example

In the following example, the system loops through all key figures of the planning level using the KeyFigure variable. Key figures, which contain the value 0, are assigned the value 1. The other key figures remain unchanged.

DATA KeyFigure TYPE KEYFIGURE_NAME.
FOREACH KeyFigure.
..IF { KeyFigure } = 0.
....{ KeyFigure } = 1.
..ENDIF.
ENDFOR.

Conditional Statement

Here you determine that a statement block is only executed if the specified condition is fulfilled. The conditional statement is always introduced by the language element IF and closed with ENDIF. You can include as many alternative conditions in a conditional statement as you like with the language element ELSEIF. As soon as one of the specified conditions is correct, the statement block belonging to it is executed, and then the statement is exited. With the language element ELSE, you can plan one statement block, which is then executed if none of the conditions specified before apply.

Syntax

IF expression1.
..statement1.
[ELSEIF expression2.
..statement2. ]
[ELSE.
..statement3. ]
ENDIF.

Explanations

expressionn:

Logical expression. If expressionn results in the value True, then the appropriate statement block statementn is executed. Otherwise the next logical expression is checked (if available), or the statement block assigned to the language element ELSE is executed (if available): Then processing is continued behind the language element ENDIF.
You can use the usual comparison operators in a logical expression (=, <>, <, >, <=, >= or EQ, NE, LT, GT, LE, GE) as well as the following additional operator:
NOT: Inverts the logical value of the subsequent expression.
IS NIL: Delivers the value True, if it was attempted to read over and above the last data record.
IS INITIAL: Delivers the value True, if a key figure or characteristic contains the initial value.
The operators AND and OR are available for the interlinkage of several logical expressions to a complex expression. You make the grouping of subexpressions within a complex expression with round brackets.

statementn:

Statement block, which is then executed, when the check of the previous logical statement results in the value True. Otherwise the statement block is skipped.

Example

In the following example sales per customer are planned. The size of the planned sales increase depends on the size of previously made sales: The sales with customers, to which a sales revenue of more than 50000 € is already allotted, should increase by 10%, while the increase with customers under this threshold should amount to 20%. The formula operand used in the example has the form { key figure name, customer, version }. Version 0 contains the actual data, version 1 the plan data.

DATA Customer TYPE 0CUSTOMER.
FOREACH Customer.
..IF { REVENUE, Customer, 0 } >= 50000.
.... { REVENUE, Customer, 1 } = { REVENUE, Customer, 0 } * 1.1.
..IF { REVENUE, Customer, 0 } >= 50000.
.... { REVENUE, Customer, 1 } = { REVENUE, Customer, 0 } * 1.2.
..ENDIF.
ENDFOR.

Loop Constructs

The formula extensions offer two different language elements for the programming of loops: DO ... ENDDO and FOREACH ... ENDFOR. Both language elements are used for different purposes:

The statement DO ... ENDDO executes a statement block between the language elements until a termination condition is fulfilled. When this condition is fulfilled, processing must be explicitly terminated with the command EXIT. Processing is continued in this case with the statement which follows ENDDO.

The statement FOREACH ... ENDFOR executes an iteration over the values of a characteristic sorted in ascending order (restricted by selection conditions from planning level or package, if required) or over the key figures contained in the planning level. The reference to the characteristic or list of key figures must always take place via a previously declared local variable. The termination criterion is implicitly specified when the last characteristic value or last key figure is reached.

Syntax

1. DO ... ENDDO

DO.
..statement1.
..IF expression1.
.. .[statement2.]
.... EXIT.
..ENDIF.
ENDDO.

2. FOREACH ... ENDFOR

FOREACH varname [, varname].
..statement1.
ENDFOR.

Explanations

expression1:

Logical expression, which defines the termination criterion for loop processing. When expression1 results in the value True, then the appropriate statement block statement2 is executed (if available). In any case, the EXIT command must be used when the termination condition is reached, as otherwise there is no possibility to exit the loop.

statement1:

Statement block, which is executed in every loop pass.

statement2:

Optional statement block, which is executed when the termination condition is reached.

varname:

Variable, which refers to a characteristic contained in the planning level, or to the quantity of key figures contained in the planning level (type KEYFIGURE_NAME).

Note

With the FOREACH loop you can not only iterate over the values of an individual characteristic, but also over characteristic combinations, which are contained in the transaction data of the package. For this you should include the characteristics involved in a list separated by commas in the FOREACH statement (for example FOREACH var1, var2, var3.). In this way, in most cases, you achieve a considerably quicker processing than with several FOREACH statements nested in each other, which only iterate over one of the characteristics.

Example

1. DO ... ENDDO

The example shows how you can calculate the factorial of a number (here the factorial of 5) with the statement DO ... ENDDO. ENDDO the factorial of a number (here the factorial of 5) can be calculated. It is very important that the loop passes are counted in the variable Counter, and the loop is exited with the EXIT statement when the termination criterion is reached.

DATA Factorial TYPE I.
DATA Counter TYPE I.
DATA Result TYPE I.

Result = 1.
Counter = 0.
Factorial = 5.

DO.
..Counter = Counter + 1.
..IF Counter <= Factorial.
....Result = Result * Counter.
..ELSE.
....EXIT.
..ENDIF.
ENDDO.

2. FOREACH ... ENDFOR

In the following example, the system loops through all key figures of the planning level. Key figures, which contain the value 0, are assigned the value 1. The other key figures remain unchanged.

DATA KeyFigure TYPE KEYFIGURE_NAME.
FOREACH KeyFigure.
..IF { KeyFigure } = 0.
....{ KeyFigure } = 1.
..ENDIF.
ENDFOR.

Messages

You can issue messages to point out errors and special features to users when executing a formula function, or to give information about processing progress with time-consuming operations.

Syntax

MESSAGE Tnnn (class) [WITH var1 [[[var2] var3] var4]].

Explanations

Type of message. Allowed values are E (error) and I (information).

Note

If you issue a message of type E, then the results of the planning function are not written in the internal transaction data buffer, and can therefore not be posted (see Saving Concept).

nnn:

Three-figure number of the message to be issued within the message class.

class:

Entry of the message class enclosed in round brackets, to which the message to be issued belongs.

Recommendation

As far as possible, use messages from a message class you created in the ABAP Workbench. In this way, you make sure that your messages are independent of possible changes to delivered message classes. However, you must transport the message class you created, from the test system into the production system.

WITH var1 [[[var2] var3] var4]:

Optional extra, to issue up to four text variables in the message text. You can use local variables, which reference characteristics or the list of key figures, as text variables, or text constants. The text variables are inserted into the static message text (short and long text) in the places, which were prepared for it with the placeholder &.

Example

In this example, the planned price of an article is increased by 10% for all customers. If the price for a customer is with the value 0 in the data basis, then an error message is issued, which also contains the name of the customer. Because of the system behavior described above with messages of the type E, no key figure values at all are changed in this case (also not for data records in which the price has a value different to 0). To execute the price increase, you must first change the data records with the value 0 or remove it from the selection.

The formula operand used in the example has the form { key figure name, customer, version }. Version 0 contains the actual data, version 1 the plan data.

DATA Customer TYPE 0CUSTOMER.
FOREACH Customer.
..IF { Price, Customer, 0 } = 0.
....MESSAGE E001(/MSG01) WITH Customer.
..ELSE.
... { Price, Customer, 1 } = { Price, Customer, 0 } * 1.1.
..ENDIF.
ENDFOR.

Special Data Access Functions

Access to Time Characteristics

You can access time characteristics with the function TMVL. The function returns the current value of the specified time characteristic, if necessary, marked up with an offset, which is also specified. The entry of an offset offers the advantage that in this way such characteristics can also be delivered, which are not contained in the transaction data of the current planning package.

Syntax

TMVL(characteristic name, offset)

Explanations

characteristic name:

Technical name of the time characteristic, whose value should be calculated.

offset:

Whole number offset, with which the current value of the time characteristic is marked up. The entry of the offset is obligatory. If you want to transfer the value of the time characteristic unchanged, then enter an offset of 0.

Example

In the following example, the net book values of an asset are calculated with straight-line depreciation over a five year duration of depreciation. Cost price is 1000, net book value 100, depreciation percentage rate 20%. The function TMVL determines the value of the time characteristic FYEAR and adds an offset of 1. The value increased by 1, is assigned to FYEAR within the DO...ENDDO loop, so that the calculation of the net book value is executed for a sequence of five fiscal years. After that the iteration is terminated. The operand for this formula has the form { key figure name, fiscal year }.

DATA YEAR TYPE I.
DATA FYEAR TYPE 0FISCYEAR.
FYEAR = VARV('ACTYEAR').
DO:
YEAR = YEAR + 1.
. FYEAR = TMVL(FYEAR, 1).
. {0AMOUNT, FYEAR} = DECL(1000, 100, 20, YEAR}.
. IF YEAR = 5.
. . EXIT.
. ENDIF.
ENDDO.

Access to Attributes

With this function, you can access data, which neither concerns characteristics or key figures, but master data attributes, which can be assigned to a characteristic.

Note

The attribute value of a characteristic that is compounded to a superordinate characteristic, can only be determined with the ATRV function, if the superordinate characteristic is not included in the list of characteristics to be changed.

Syntax

ATRV(attributname, varname)

Explanations

attributname:

Technical name of the attribute, whose value should be calculated.

varname:

Variable, which refers to the characteristic, to which the attribute belongs.

Example

In this example, the prices are conducted as attributes for the characteristic article. You use the ATRV function to determine the attribute value. The function expects the technical name of the attribute to be read (for example 0PRICE) as parameter, and a variable, which refers to the characteristic, to which the attribute belongs (for example 0ARTICLE). The operand for this formula has the form { key figure name, article }.

DATA ARTICLE TYPE 0ARTICLE.
FOREACH ARTICLE.
. {REVENUE, ARTICLE} = ATRV('0PRICE', ARTICLE) * {QUANTITY, ARTICLE}.
ENDFOR.

Access to Variables

With this function, you can determine the value of a (global) Variable, which was defined in the context of the current planning area.

Note

In the context of formula extensions, only such variables can be used, which are restricted to single value. Otherwise the system cannot clearly determine the value of the variables.

Direct access to a global variable using its name (in a value assignment with the form "{Operator} = varname.") is not possible. Always use the VARV function for this purpose.

Syntax

VARV(varname)

Explanations

varname:

Technical name of the variable.

Example

In the following example two variables of the type 0CUSTOMER are declared. One of them is used as a control variable for an iteration over all customers. The other one is reserved using the VARV function with the value of the global variable VCUST, as it was defined in the context of the planning area. In the iteration, as soon as the customer defined by the variable is reached, the planned revenue for all articles is increased by 10%. Subsequently, the iteration is terminated with the EXIT statement. The operand for this formula has the form { key figure name, article }.

DATA ARTICLE TYPE 0ARTICLE.
DATA CUSTOMER TYPE 0CUSTOMER.
DATA VCUSTOMER TYPE 0CUSTOMER.

VCUSTOMER = VARV('VCUST').
FOREACH CUSTOMER.
. IF CUSTOMER = VCUSTOMER.
. . FOREACH ARTICLE.
. . . {REVENUE, ARTICLE} = (REVENUE, ARTICLE) * 1.1.
.. ENDFOR.
. . EXIT.
. ENDIF
ENDFOR.

Determination of the Current Characteristic Value

With this function, you can transfer the current value of a characteristic to a variable, and in this way find out which data was currently edited. The function OBJV is called up without entering parameters. First the type of the variable (and with it the characteristic, to which the variable refers) is determined, to which the function result should be assigned. Then the function determines the current value of the characteristic referenced by the variable, and returns the value as function result.

Above all, the function is used to keep the list of fields to be changed as small as possible. This has the following advantages:

The complexity of the formula operands decreases.
Iterations, which are superfluous in content, but syntactically necessary can be reduced using characteristic values. The performance improves through this.

Note

The result of the function OBJV can only be assigned to such variables, which were declared with reference to one of the characteristics, which are not contained in the list of fields to changed of the formula function.

Syntax

varname = OBJV()

Explanations

varname:

Technical name of the variable.

Example

In the following example, the example used above for the VARV function is enhanced in the following way: An additional variable is declared of the type 0COUNTRY, to which the current value of the characteristic is transferred within the FOREACH loop. Revenue planning is executed as in the example above, however, not for articles that were sold in Germany. Prerequisite for calling up the OBJV() function is that the characteristic 0COUNTRY is not contained in the quantity of fields to be changed.

DATA ARTICLE TYPE 0ARTICLE.
DATA COUNTRY TYPE 0COUNTRY.
DATA CUSTOMER TYPE 0CUSTOMER.
DATA VCUSTOMER TYPE 0CUSTOMER.

VCUSTOMER = VARV('VCUST').
FOREACH CUSTOMER.
. IF CUSTOMER = VCUSTOMER.
. . FOREACH ARTICLE.
. . . COUNTRY = OBJV().
. . . IF COUNTRY = 'DE'.
. . . . {REVENUE, ARTICLE} = (REVENUE, ARTICLE) * 1.1.
. . . ENDIF.
.. ENDFOR.
. . EXIT.
. ENDIF
ENDFOR.

See also:

Formula Extension: Examples