Approximate numeric data types

Adaptive Server Enterprise differs from SQL Anywhere and Sybase IQ in how the FLOAT(precision) data type is interpreted: that is, when to create a 4-byte data type and when to create an 8-byte data type.

REAL values are four bytes, DOUBLE values are eight bytes. Turning on the FLOAT_AS_DOUBLE option makes the FLOAT keyword behave like Adaptive Server Enterprise's FLOAT keyword when a precision is not specified. Since Adaptive Server Enterprise treats its own FLOAT values as DOUBLE, enabling the option makes Sybase IQ to treat FLOAT values in the same way Adaptive Server Enterprise treats FLOAT values.

Turning off the FLOAT_AS_DOUBLE option makes the FLOAT keyword behave like a Sybase IQ and SQL Anywhere four byte REAL value.

For Adaptive Server Enterprise, the precision specified in FLOAT(precision) means decimal precision. For SQL Anywhere and Sybase IQ, precision is an integer expression that specifies the number of bits used in the mantissa of the floating point number.

For more information on the Sybase IQ FLOAT data type, see Numeric data types in Chapter 3, “SQL Data Types.”