Mathematical Functions 

ABAP contains a range of built-in functions that you can use as mathematical expressions, or as part of a mathematical expression:

[COMPUTE] <n> = <func>( <m> ).

The blanks between the parentheses and the argument <m> are obligatory. The result of calling the function <func> with the argument <m> is assigned to <n>.

Functions for all Numeric Data Types

The following built-in functions work with all three numeric data types (F, I, and P) as arguments.

Functions for all numeric data types

Function

Result

ABS

Absolute value of argument.

SIGN

Sign of argument:                      1 X > 0

                                              SIGN( X) = 0 if X = 0

                                                               -1 X < 0

CEIL

Smallest integer value not smaller than the argument.

FLOOR

Largest integer value not larger than the argument.

TRUNC

Integer part of argument.

FRAC

Fraction part of argument.

The argument of these functions does not have to be a numeric data type. If you choose another type, it is converted to a numeric type. For performance reasons, however, you should use the correct type whenever possible. The functions itself do not have a data type of their own. They do not change the numerical precision of a numerical operation.

DATA N TYPE P DECIMALS 2.
DATA M TYPE P DECIMALS 2 VALUE '-5.55'.

N = ABS( M ).   WRITE:   'ABS:  ', N.
N = SIGN( M ).  WRITE: / 'SIGN: ', N.
N = CEIL( M ).  WRITE: / 'CEIL: ', N.
N = FLOOR( M ). WRITE: / 'FLOOR:', N.
N = TRUNC( M ). WRITE: / 'TRUNC:', N.
N = FRAC( M ).  WRITE: / 'FRAC: ', N.

The output appears as follows:

ABS:              5.55

SIGN:             1.00-

CEIL:             5.00-

FLOOR:            6.00-

TRUNC:            5.00-

FRAC:             0.55-

DATA: T1(10),
T2(10) VALUE '-100'.

T1 = ABS( T2 ).

WRITE T1.

This produces the following output:

100

Two conversions are performed. First, the contents of field T2 (type C) are converted to type P. Then the system processes the ABS function using the results of the conversion. Then, during the assignment to the type C field T1, the result of the function is converted back to type C.

Floating-Point Functions

The following built-in functions work with floating point numbers (data type F) as an argument.

Functions for floating point data types

Function

Meaning

ACOS, ASIN, ATAN; COS, SIN, TAN

Trigonometric functions.

COSH, SINH, TANH

Hyperbolic functions.

EXP

Exponential function with base e (e=2.7182818285).

LOG

Natural logarithm with base e.

LOG10

Logarithm with base 10.

SQRT

Square root.

For all functions, the normal mathematical constraints apply (for example, square root is only possible for positive numbers). If you fail to observe them, a runtime error occurs.

The argument of these functions does not have to be a floating point field. If you choose another type, it is converted to type F. The functions themselves have the data type F. This can change the numerical precision of a numerical operation.

DATA: RESULT TYPE F,
PI(10) VALUE '3.141592654'.

RESULT = COS( PI ).

WRITE RESULT.

The output is -1.00000000000000E+00 . The character field PI is automatically converted to a type F field before the calculation is performed.