FLOAT

An item of type FLOAT is interpreted as a binary value for double-precision floating-point numbers with as many as 16 significant digits. The length is fixed at 8 bytes. In EGL-generated Java programs, the value ranges from 4.9e-324 to 1.7976931348623157e308.

FLOAT corresponds to each of these definitions:

• The FLOAT data type in a relational database management system

• The double data type in C, C++, or Java

For floating-point values, format conversion between Java and host COBOL formats is supported by DB2 but is not supported on calls to host programs.

Related reference
Primitive types