Fixed Decimal declarations:

    type of data: coded arithmetic (Binary Coded Decimal, packed two decimal digits to a single 8-bit byte)

    S/360, S/370 data format: packed decimal

    default precision: five (5) decimal digits

    maximum precision: 15 decimal digits

    example:

    DECLARE DOLLAR_AMOUNT FIXED DECIMAL (9,2) INIT (100.00);

    First number in declaration is the total number of stored digits. Second number in declaration (optional) is the number of digits to the right of an implied decimal point. If the second number is zero or left out, then the number is an integer.

    Fixed decimal is more efficient if the precision is declared as an odd number of digits (because of extra code generated to guarantee that the last half of an even number of decimal digits remains zero). Numbers too large to fit into the specified storage will be truncated. Numbers too small to fill the entire storage space will be zero-filled.

    Fixed Binary declarations:

    type of data: coded arithmetic (binary )

    S/360, S/370 data format: fixed point

    default precision: 16 bits

    maximum precision: 31 bits

    example:

    DECLARE LOCAL_DISTANCE FIXED BINARY (31,16) INIT (256.5);

    First number in declaration is the total number of stored bits. Second number in declaration (optional) is the number of bits to the right of an implied “decimal” point. If the second number is zero or left out, then the number is an integer.

    Generally the fastest execution time for arithmetic. Sizes 16 bits or fewer are stored in two bytes. Sizes 17-31 bits are stored in four bytes.

    Undeclared variables starting with the letters I through N, inclusive, default to FIXED BINARY (15).

    Many decimal fractions have no exact binary equivalent, which may result in a small rounding error.