The class CORE::BigFloat is a variable precision floating-point type. Rounding mode and precision (i.e. mantissa length) of CORE::BigFloat can be set. Since it also carries the error of a computed value.
This number type is provided by the Core library [KLPY99].
Cgal defines the necessary functions so that this class complies to the requirements on number types.